betaFeedback

WorksPhysics and Astronomy

Solid-state reaction synthesis of superconducting refractory-metal germanides for Ge/SiGe quantum devices

Abstract

Solid-phase reactions offer a route to integrate superconductors into Ge/SiGe heterostructures, but forming high-quality superconducting layers near the buried Ge quantum well remains challenging. We investigate reactions of V, Ta, and Nb with elemental Ge and with the Si0.2_{0.2}Ge0.8_{0.8} barrier of Ge/SiGe heterostructures at processing temperatures of 365-390 ∘^\circC. Cross-sectional electron microscopy and low-temperature electrical transport show that deposition on Ge at 390 ∘^\circC, in which approximately 2 nm metal increments are separated by growth stops for in-situ annealing, produces reactions throughout the deposited films and superconducting transitions at approximately 0.14 K for V-Ge, 2.1 K for Ta-Ge, and 2.8 K for Nb-Ge. The Nb-Ge layer has an out-of-plane critical field of approximately 0.8 T at 2.1 K. In the heterostructure samples, intermixing is confined to regions approximately 5-12 nm thick near the metal-SiGe interface, leaving a substantial residual metal layer. Microscopy and transport together suggest that this residual film dominates the measured superconducting response. These results establish low-temperature formation of superconducting refractory-metal-Ge layers while identifying the limited interfacial reaction as the principal obstacle to extending this approach toward the buried quantum well.

Introduction

Planar Ge/SiGe heterostructures have emerged as a promising platform for hole-spin qubits [1, 2, 3, 4, 5]. Holes in strained Ge combine large out-of-plane gg factors (∣gz∣≈10|g_z| \approx10–20) [6, 7], strong intrinsic spin–orbit coupling that enables electrical qubit control [2, 8], mobilities exceeding 7×106 cm2 V−1s−17 \times10^{6}\ \mathrm{cm}^{2}\ \mathrm{V}^{-1}\mathrm{s}^{-1} [9, 10, 11, 12, 13], and a nuclear-spin-poor lattice that can be isotopically purified [14, 15, 16]. These properties make Ge attractive not only for isolated spin qubits, but also for hybrid architectures in which semiconductor quantum devices are integrated with superconducting elements.

Superconductivity could extend the Ge platform in two complementary directions. Superconducting resonators can mediate long-range microwave coupling between spin qubits, as demonstrated in Si [17, 18, 19, 20, 21, 22] and now pursued in Ge [23, 24]. Proximity-induced pairing in the hole gas provides a route to Josephson devices, proximitized quantum dots, Andreev spin qubits, quantum-dot Kitaev chains, and topological superconductivity [25, 26, 27, 28, 29, 30, 31, 32, 33]. In such devices, the semiconductor–superconductor interface is a central material parameter, as it sets the strength and quality of the induced superconductivity [34].

Most superconducting Ge devices so far have relied on metal deposition onto an etched SiGe top barrier [35, 36, 37]. Etching can damage the semiconductor surface and reduce interface transparency. A different strategy is to form the superconductor directly inside the heterostructure by a solid-phase reaction. Tosato et al. demonstrated this approach with PtSiGe formed inside the SiGe barrier at 400°C, obtaining a hard induced gap [38]. PtSiGe has since been reproduced independently in planar Ge Josephson junctions with a comparable induced gap [39] and has enabled a proximitized quantum dot [33]. PtSiGe itself, however, has a relatively low critical temperature, Tc≈0.5T_\mathrm{c} \approx0.5 K, and an in-plane critical field of approximately 0.4 T, limiting the accessible energy and magnetic-field scales. Partial substitution of Ir for Pt extends these scales, yielding a PtIrGe alloy with Tc≈1.85T_\mathrm{c} \approx1.85 K and an out-of-plane critical field of 1.9 T [40] at 80 mK. Solid-phase epitaxy of a broader set of group-VIII germanosilicides—including Pd-, Rh-, and Ir-based compounds—has been characterized by Li et al., who found IrGe superconducting at up to 3.4 K [41].

Refractory-metal germanides offer a further route to higher energy and magnetic-field scales. V, Ta, and Nb form several superconducting germanide and silicide phases [42, 43], including the A15 compounds Nb3_3Ge and V3_3Si, which belong to the same structural family as Nb3_3Sn and have critical temperatures up to approximately 23 K and high upper critical fields [44, 45, 46, 47, 48]. These phases are normally synthesized at temperatures above 700°C or from stoichiometric sputter targets [47, 49]. In strained Ge/SiGe heterostructures, by contrast, the quantum well and top barrier are grown near 500°C [11], which places a practical upper bound on subsequent thermal processing if the quantum well is to be preserved. It is therefore not known whether useful superconducting refractory-metal germanides can be formed within the heterostructure thermal budget, or whether these metals can react sufficiently far into the SiGe barrier to approach the buried quantum well.

Here we address these two questions by comparing V, Ta, and Nb deposited at 365–390°C onto bulk Ge and onto the Si0.2Ge0.8\mathrm{Si}_{0.2}\mathrm{Ge}_{0.8} top barrier of Ge/SiGe heterostructures. We combine cross-sectional scanning transmission electron microscopy with low-temperature resistance measurement. On Ge, all three metals react through the deposited layer and form superconducting germanides, with Ta–Ge and Nb–Ge reaching Tc=2.1T_c = 2.1 and 2.82.8 K, respectively. On the heterostructures, the same thermal budget produces only a thin reacted region at the metal–SiGe interface. The contrast shows that low-temperature germanide formation is feasible, but that reaction kinetics in the SiGe barrier currently limit extension of the process toward the quantum well.

Experimental methods

Heterostructures and substrates

The heterostructures, grown on a Si substrate, consist of a strained Ge quantum well (16 nm) buried under a 55 nm thick Si0.2Ge0.8\mathrm{Si}_{0.2}\mathrm{Ge}_{0.8} top barrier [9]. Since the barriers and quantum well are grown at 500°C [11], we use this temperature as the upper thermal budget for all subsequent processing. Samples of bulk undoped Ge (001) wafers were processed in parallel to separate the metal–Ge reaction from the additional chemical and structural complexity of the SiGe barrier.

Surface preparation

Native GeOx\mathrm{GeO}_x was removed from the Ge samples in deionized water at 90°C for 10 min [50]. The native SiOx\mathrm{SiO}_x on the heterostructures was etched in HF:H2_2O (1:7) for 30 s, after which the samples were transferred under isopropanol. Both substrate types were then dried with N2\mathrm{N}_2 and loaded into the evaporator load lock within 1–2 min.

Metal deposition and annealing

V, Ta, and Nb films with total thicknesses of 30–40 nm were deposited by electron-beam evaporation in an ultra-high-vacuum chamber with a pressure of 10−910^{-9}–10−710^{-7} mbar. The substrates were kept at a fixed temperature using a heater directly below the sample, regulated by a thermocouple, with a maximum temperature of 390°C. Most films were deposited step-wise: approximately 2 nm of metal was deposited at 0.05–0.1 Å s−1^{-1} and followed by a 5–15 min anneal at the growth temperature before the next increment was added [50]. In continuous-growth samples, the full metal thickness was deposited at 0.05 Å s−1^{-1} and annealed only after deposition. Unless stated otherwise, the samples described in the main text were prepared by the step-wise process.

After the final anneal, the films were capped to suppress oxidation. Most samples received 10 nm of AlOx\mathrm{AlO}_x deposited at 300 K using e-beam evaporation. One Nb/SiGe sample instead received a 3 nm Ta cap, which can act as an oxygen-diffusion barrier [51].

Transport measurements

Unpatterned films were wire-bonded in a four-terminal in-line geometry and measured in a dilution refrigerator between 20 mK and several kelvin. A vector magnet provided fields up to 1 T perpendicular to the film and 5.5 T in plane. The resistance was measured using standard lock-in techniques and low-noise differential amplification with excitation currents of 0.1–1 μ\muA. We define TcT_c as the temperature at which the resistance falls to 0.1RN0.1R_N, where RNR_N is the normal-state resistance. The out-of-plane critical field BcB_c is defined using the same 10% criterion at fixed temperature.

Electron microscopy

Cross-sectional lamellae were prepared by focused-ion-beam milling (FIB) in a Helios G4 CX, using a Pt or W capping layer to protect the sample and placed on a Cu support. High-angle annular dark-field scanning transmission electron microscopy (HAADF-STEM) and energy-dispersive X-ray spectroscopy (EDX) were performed in a CsC_s-corrected FEI Titan Cubed TEM operated at 300 keV. The cross-sections were exposed to air before microscopy and therefore contain a surface oxide that is not part of the as-grown buried interface. Atomic compositions were quantified from EDX spectra (excluding C and Cu) and carry an uncertainty of roughly 10% for the transition metals and Ge, while for the remaining elements it is < 2%. The uncertainty has not been plotted for clarity of the data. The protective capping layer materials for FIB Pt/W and Al in the AlOx\mathrm{AlO}_x capping are not shown, but are accounted for in the atomic fraction results below.

Results

Table I summarizes the samples and their superconducting properties. Our central comparison is between reactions on elemental Ge and on the Si0.2Ge0.8\mathrm{Si}_{0.2}\mathrm{Ge}_{0.8} top barrier of the heterostructure. For each material, we first establish whether the chosen thermal budget is sufficient to form a reacted superconducting layer on Ge and then apply a comparable recipe to the heterostructure. This side-by-side approach separates the intrinsic ability of the metal to form a superconducting germanide from the additional kinetic constraints introduced by the SiGe alloy.

MetalSubstrateThickness (nm)GrowthTemperature (°C)Anneal (min)CapReacted layerTcT_{\mathrm{c}} (K)BcB_{\mathrm{c}} (T)
VGe15×215 \times2step39015×515 \times5AlOx_xcomplete, V:Ge ≈1:1\approx1:1*0.14Not measurable
VGe40cont.365–AlOx_xNA4.00.82 (2.3 K)
VSiGe40cont.365–AlOx_x≈7\approx7 nm4.150.16 (3.8 K)
TaGe15×215 \times2step39015×1515 \times15AlOx_xcomplete, Ta:Ge ≈2:3\approx2:3*2.10.19 (1.8 K)
TaSiGe15×215 \times2step39015×1515 \times15AlOx_x≈5\approx5–10 nm1.040.31 (0.8 K)
NbGe20×220 \times2step39020×520 \times5AlOx_xcomplete, Nb:Ge ≈5:4\approx5:4*2.80.77 (2.1 K)
NbSiGe15×215 \times2step39015×515 \times5Ta≈7+5\approx7 + 5 nm4.70.65 (4.0 K)

Table 1. Samples and superconducting properties. Depositions were performed at 365–390°C. Step-wise (step) samples were annealed for 5–15 min after every approximately 2 nm increment, whereas continuous (cont.) samples were annealed after deposition while cooling down to room temperature. TcT_{\mathrm{c}} is defined by R=0.1RNR = 0.1R_{\mathrm{N}}, and BcB_{\mathrm{c}} is the out-of-plane critical field at the stated temperature using the same resistance criterion. Compositions marked * require verification. Reacted layers listed as NA were not subjected to TEM.

Low-temperature formation of superconducting germanides on Ge

We first test whether V, Ta, and Nb can react with Ge at temperatures below the 500°C heterostructure growth temperature. All three metals show complete reaction through the deposited film under at least one of the processing conditions studied here, demonstrating that superconducting refractory-metal germanides can be formed at 365–390°C.

For V, the superconducting target phases include the A15 compounds V3_3Si (Tc≈17T_{\mathrm{c}} \approx17 K) [44], V3_3Ge (6.2 K) [52], and the pseudobinary V3_3Six_xGe1−x_{1-x} system [53]. These phases are normally prepared at substantially higher temperatures, and their superconducting properties are sensitive to stoichiometry and crystal structure. Figure 1(a,Figure 1b) shows a V/Ge sample prepared by depositing 15×215 \times2 nm of V step-wise at 390°C with 5 min waiting times between the steps. EDX reveals that the deposited V is incorporated throughout a reacted V–Ge layer with an approximate V:Ge ratio of 1:1. This composition does not correspond to the equilibrium compounds V3_3Ge, V5_5Ge3_3, V11_{11}Ge8_8, or V17_{17}Ge31_{31} [54], suggesting a disordered or multiphase reaction product rather than a well-defined A15 phase. The layer becomes superconducting at Tc=140T_{\mathrm{c}} = 140 mK, with nearly identical transitions measured at two positions on the sample [Fig. 1(a)]. Both the composition and the transition temperature therefore argue against formation of V3_3Ge.

Vanadium on Ge and on the Ge/SiGe heterostructure

Figure 1. Vanadium on Ge and on the Ge/SiGe heterostructure. (a) Resistance normalized to the normal-state value RNR_N as a function of temperature for V deposited step-wise on Ge (15×215 \times2 nm, 390°C, 5 min anneals), measured at two positions on the film. The transition reaches R=0.1RNR = 0.1R_N at 135 mK. The inset shows the four-terminal geometry and the definitions of B⊥B_\perp and B∥B_\parallel. (b) Cross-sectional HAADF-STEM image, EDX maps of V, Ge, and O, and corresponding atomic-fraction depth profile for the same sample. The red outline in the HAADF image marks the data used for the atomic fraction profile. V and Ge occupy the reacted layer between the AlOx_x cap and the Ge substrate. Residual oxygen near the original Ge surface marks the approximate initial interface, while voids appear immediately below it on the Ge side. (c) R/RNR/R_N as a function of temperature for 40 nm V deposited continuously at 365°C on the Ge/SiGe heterostructure and, using the same deposition conditions, on Ge at two positions. All three transitions occur between 4.0 and 4.15 K. (d) HAADF-STEM image, EDX maps of V, Ge, Si, and O, and depth profile of the heterostructure sample. The V film intermixes with the SiGe barrier only over approximately 7 nm near the interface.

The V/Ge cross-section also provides information about the reaction mechanism. A residual oxygen signal remains near the depth of the original Ge surface, while voids are visible directly below this marker plane on the Ge side [Fig. 1(b)]. We interpret this combination as evidence for a Kirkendall-type reaction in which Ge is the dominant diffusing species. In this picture, Ge diffuses into the V layer faster than V diffuses into the substrate, producing a net vacancy flux toward the Ge side of the original interface where vacancies can condense into voids. A similar diffusion asymmetry is known for refractory-metal disilicides [55, 56]. Etch-induced pits cannot be excluded entirely, but their coincidence with the marker plane would be less natural.

Ta provides a second test of the low-temperature reaction route. Relevant superconducting phases include TaSi2_2 (Tc=0.35T_{\mathrm{c}} = 0.35 K) [57], TaGe2_2 (approximately 1.9–2.7 K depending on growth method) [43, 50, 58], and metastable A15 Ta3_3Ge (Tc≈8T_{\mathrm{c}} \approx8 K) [59]. Following the step-wise approach of Stassen et al. [50], we deposited 15×215 \times2 nm of Ta on Ge. EDX indicates that the Ta is fully incorporated into a reacted layer [Fig. 2(b)]. The current quantification gives a Ta:Ge ratio of approximately 2:3, which lies between the equilibrium compounds Ta5_5Ge3_3 and TaGe2_2 [54]. This ratio depends on the chosen X-ray lines, since the Ta-L lines overlap with the Ge-K lines; we therefore quantified the atomic fractions using the Ta-M and Ge-K lines. More generally, the reported composition of Ta–Ge films depends strongly on the measurement technique. For Ta deposited on Ge(001) at 400°C, Stroobeen et al. obtained approximately 38.5% Ta from EDS, close to our value, whereas Rutherford backscattering spectrometry (RBS) on the same film gave a Ta-rich Ta:Ge ratio of 58.5:41.5, and XPS depth profiling likewise indicated a Ta-rich film [50]. They attributed the Ta deficit in EDS to oxidation during the sample transfer, overlapping line shapes, and the absence of a suitable standard, adopted the RBS value, and interpreted it using the phase diagram as a mixture of predominantly Ta5_5Ge3_3 with some TaGe2_2 [50]. Their films showed no Ta–Ge diffraction peaks, and our TEM likewise did not identify a crystalline phase in the Ta or Nb samples, so in neither case can the phase be assigned structurally. The discrepancy of about 20 percentage points between EDS and RBS in Ref. [50] exceeds our quoted EDX uncertainty, and our 2:3 ratio may therefore underestimate the Ta content. A standard-free technique such as RBS, or atom probe tomography, would be needed to establish the absolute composition.

The Ta–Ge layer becomes superconducting at Tc=2.1T_{\mathrm{c}} = 2.1 K [Fig. 2(a)]. At 1.8 K, superconductivity persists to an out-of-plane field of order 0.19 T using the present resistance criterion. The transition temperature is consistent with reported TaGe2_2 films and crystals [50, 58], although the present microscopy does not yet establish that this is the only phase present.

Among the three materials, Nb gives the most favourable superconducting properties on Ge. Nb forms high-TcT_c A15 compounds Nb3_3(Ge,Si) [45, 49], as well as lower-temperature phases such as chiral C40 NbGe2_2 (Tc=2.06T_c = 2.06 K) [60] and NbSi2_2 (Tc=0.13T_c = 0.13 K) [61]. Because the A15 phases generally require much higher formation temperatures, the low-temperature process is more naturally expected to produce a non-A15 germanide. Step-wise deposition of 20×220 \times2 nm Nb on Ge produces a fully reacted layer [Fig. 3(b)] with an approximate Nb:Ge ratio of 5:4. This composition lies between the equilibrium compounds Nb5_5Ge3_3 and NbGe2_2 [54], and may therefore represent a disordered or multiphase layer rather than a single equilibrium phase.

The reacted Nb–Ge layer has Tc=2.8T_c = 2.8 K [Fig. 3(a)], exceeding the reported bulk transition temperature of NbGe2_2 [60]. At 2.1 K, the layer remains superconducting to an out-of-plane field of approximately 0.77 T, the largest field scale observed among the reacted Ge samples. Taken together, the V, Ta, and Nb results establish that the 365–390°C processing window is sufficient to form superconducting refractory-metal germanides on elemental Ge, even though the resulting compositions need not correspond to the equilibrium high-TcT_c phases.

Reaction is confined near the SiGe interface

Next, we apply comparable processing conditions to the Si0.2_{0.2}Ge0.8_{0.8} top barrier. In all three material systems, the result differs qualitatively from the corresponding Ge sample: reaction is confined to a narrow interfacial region and a substantial fraction of the deposited metal remains unreacted. The transport response must therefore be interpreted together with the microscopy rather than assigned directly to the reacted layer.

For V, 40 nm deposited continuously at 365°C intermixes with the SiGe barrier only within approximately 7 nm of the interface Figure 1(d). The film shows a transition at Tc=4.15T_{\mathrm{c}} = 4.15 K Figure 1(c), which is more likely to be a remaining metallic V film, with a modest suppression that could arise from reduced thickness or disorder [62]. A V film deposited continuously on Ge using the same 365°C parameters shows nearly the same transition at Tc=4.0T_{\mathrm{c}} = 4.0 K Figure 1(c), suggesting that continuous growth followed by a single anneal is insufficient to drive the strong reaction observed for the step-wise V/Ge sample. The lower growth temperature used for the continuous films may also contribute. However, this sample was not examined by TEM, so we cannot distinguish an essentially unreacted V film from a different reaction product with a similar TcT_{\mathrm{c}}, such as a more V-rich germanide. Because the fully reacted V–Ge layer itself has a much lower TcT_{\mathrm{c}}, we did not pursue V as the leading candidate for proximity devices.

For Ta, the step-wise process that reacts throughout the deposited film on Ge produces only a 5–10 nm intermixed layer on the heterostructure Figure 2(d). This region remains far above the quantum well. The sample has Tc=1.04T_{\mathrm{c}} = 1.04 K and Bc=310B_{\mathrm{c}} = 310 mT at 800 mK [Fig. 2(c)]. These values are closer to the superconducting properties of disordered β\beta-Ta, for which TcT_{\mathrm{c}} is approximately 0.5 K [63], than to those of α\alpha-Ta, which has Tc≈4.4T_{\mathrm{c}} \approx4.4 K [64]. We therefore attribute the measured transport properties primarily to the remaining Ta film.

Tantalum on Ge and on the heterostructure

Figure 2. Tantalum on Ge and on the heterostructure. (a) Resistance as a function of out-of-plane magnetic field B⊥B_{\perp} and set temperature TsetT_{\mathrm{set}} for Ta deposited step-wise on Ge (15×215 \times2 nm). At zero field, the resistance transition occurs at 2.1 K. (b) Cross-sectional HAADF-STEM image, EDX maps of Ta, Ge, and O, and the corresponding depth profile. The deposited Ta is incorporated into an approximately 40 nm reacted region below the AlOx_x cap. (c) Same measurement as in (a) for the corresponding Ta process on the Ge/SiGe heterostructure. The zero-field transition is at 1.04 K and superconductivity persists to approximately 0.31 T at 0.8 K. (d) HAADF-STEM image, EDX maps of Ta, Ge, Si, and O, and the corresponding depth profile for the heterostructure sample. Most of the Ta remains in the metallic film, while only a 5–10 nm region near the interface is intermixed with the barrier. The Ge quantum well appears as a Si-free, Ge-rich band approximately 60 nm below the interface. Both samples carry an AlOx_x cap.

Niobium on Ge and on the heterostructure

Figure 3. Niobium on Ge and on the heterostructure. (a) Resistance as a function of B⊥B_{\perp} and TsetT_{\mathrm{set}} for Nb deposited step-wise on Ge (20 × 2 nm, AlOx_x cap). The zero-field transition occurs at 2.8 K and the superconducting region extends to approximately 0.77 T near 2.1 K. (b) Cross-sectional HAADF-STEM image, EDX maps of Nb, Ge, and O, and the corresponding depth profile. The deposited Nb is incorporated into an approximately 50 nm reacted Nb–Ge layer beneath the AlOx_x cap. (c) Same transport measurement for Nb deposited step-wise on the Ge/SiGe heterostructure (15 × 2 nm, Ta cap). (d) Cross-sectional HAADF-STEM image, EDX maps of Nb, Ge, Si, and O, and the corresponding depth profile for the heterostructure sample. Beneath the Ta cap, the Nb film reacts with the SiGe barrier over approximately 12 nm, forming an approximately 7 nm Nb–Ge-rich region followed by an approximately 5 nm Nb–Si–Ge interfacial region.

Nb shows the deepest intermixing of the three metals. In the Nb/SiGe sample with a Ta cap, EDX resolves approximately 7 nm of Nb–Ge-rich material above an approximately 5 nm Nb–Si–Ge region [Fig. 3(d)]. This compositional plateau followed by a transition region is distinct from the simpler interfacial profiles observed for V and Ta. The sample has Tc=4.7T_{\mathrm{c}} = 4.7 K and Bc=0.65B_{\mathrm{c}} = 0.65 T at 4 K [Fig. 3(c)]. The TEM shows that roughly 20 nm of Nb remains above approximately 12 nm of reacted Nb–(Si)Ge. The most conservative interpretation is that the observed superconductivity is dominated by the remaining Nb film, possibly modified by proximity coupling to the reacted layer [65]. A spatial distribution of local transition temperatures could further broaden the measured transition. We don’t expect thin Ta cap to contribute to superconducting properties of the measured film. Critical temperature of α\alpha-Ta is around 4.4K for crystalline thick films grows at high temperatures. Despite the material-dependent details, the common result is that the reaction stalls near the metal–SiGe interface.

Discussion

The side-by-side comparison isolates the main limitation of the materials. At 365–390°C, V, Ta, and Nb can all react extensively with elemental Ge, yet the same processes penetrate only a few nanometers into Si0.2_{0.2}Ge0.8_{0.8}. The difference is therefore not simply whether a given metal is capable of forming a germanide at this temperature; it is the kinetics of reaction with the alloy barrier. Several mechanisms could contribute. The Ge wafers and epitaxial SiGe\mathrm{SiGe} barriers are both crystalline, but their defect densities are unlikely to be identical. A higher density of defects in the bulk Ge wafers could accelerate diffusion relative to the high-quality epitaxial barrier. The SiGe\mathrm{SiGe} alloy also changes the bonding environment relative to elemental Ge, and forming a reaction product requires redistribution of both Si and Ge. Thin-film formation temperatures provide an empirical indication of the corresponding kinetic barriers: VSi2\mathrm{VSi}_2 on Si forms only near 570–650°C with an activation energy of approximately 1.7 eV, while V3Si\mathrm{V}_3\mathrm{Si} forms above approximately 730°C [66]; V, Nb, and Ta germanides likewise require comparatively high temperatures in conventional thin-film reactions [67].

Strain and point-defect physics may further modify the reaction. Kringhøj et al. showed that vacancy-mediated Sb diffusion in Si0.91Ge0.09\mathrm{Si}_{0.91}\mathrm{Ge}_{0.09} is enhanced by compressive strain and retarded by tensile strain [68]. Self-diffusion in Ge is vacancy-mediated, whereas both vacancies and interstitials can contribute in Si [69, 70]. The diffusivities of V, Ta, and Nb in Si0.2Ge0.8\mathrm{Si}_{0.2}\mathrm{Ge}_{0.8} are therefore unlikely to follow a simple interpolation between pure Si and pure Ge and require dedicated study.

The V/Ge microstructure adds another constraint on how far the reaction should ultimately be driven. The voids located on the Ge side of the apparent marker plane are consistent with Ge being the dominant moving species. If the same mechanism operates in the SiGe\mathrm{SiGe} barrier, extending the reacted layer toward the quantum well would inject vacancies into the semiconductor side of the interface. The reaction should therefore be advanced only as far as required to establish strong superconducting coupling, because excessive reaction could introduce vacancy-related disorder near the quantum well.

The comparison with PtSiGe is particularly informative [38]. Firstly, Pt was deposited at room temperature and subsequently reacted by rapid thermal annealing at 400°C, whereas in the present work V, Ta, and Nb were deposited directly onto hot substrates to promote intermixing. The annealing method may itself play a role: rapid thermal annealing is carried out in an inert-gas ambient, such as Ar [40], with fast heating and cooling ramps, whereas our films were annealed in situ in ultra-high vacuum. Differences in ambient, thermal history, and temperature calibration between the two approaches could therefore affect the reaction. Secondly, the thin-film formation temperature of PtGe is around 380–440°C [67], and PtSiGe can be formed near 400°C [38], while refractory-metal silicides and germanides generally require substantially higher temperatures. Therefore, the simplest interpretation is that V, Ta, and Nb diffuse too slowly through the SiGe barrier on the timescales used in this work. The barrier thickness may compound this difference: the 55 nm top barrier used here is more than twice as thick as the 22 nm barrier in Ref. [38].

These observations suggest three practical routes forward. First, the processing temperature can be raised to the 500°C heterostructure growth temperature: PtIrSiGe formed by rapid thermal annealing at 500°C proximizes the buried Ge quantum well, and TEM shows that the quantum well remains structurally intact [40]. Second, substantially longer post-deposition anneals may increase the reaction depth without requiring a larger metal thickness; at present, adding more metal mainly increases the amount of unreacted film. Third, a thinner top barrier would reduce the distance between the reacted region and the quantum well, at the possible cost of reduced mobility.

Conclusion

Step-wise deposition at 390°C produces superconducting reaction layers throughout the deposited V, Ta, and Nb films on elemental Ge. The resulting transition temperatures are approximately 0.14 K for V–Ge, 2.1 K for Ta–Ge, and 2.8 K for Nb–Ge. The Nb–Ge layer has an out-of-plane critical field of approximately 0.8 T at 2.1 K.

In the Ge/SiGe samples, intermixing is confined to regions approximately 5–12 nm thick near the metal–SiGe interface. Microscopy and transport suggest that the residual metal films dominate the measured superconductivity. The results identify limited interfacial reaction as the principal processing obstacle, without establishing a unique microscopic cause. They separate two requirements for Ge/SiGe integration: low-temperature formation of superconducting reaction layers, demonstrated here on elemental Ge, and controlled extension of those layers toward the buried quantum well while preserving its quality, which remains unresolved under the conditions investigated.

Acknowledgments

For the Titan TEM results, we acknowledge support from the Kavli Institute of Nanoscience, Delft University of Technology, and the Netherlands Electron Microscopy Infrastructure (NEMI), project number 184.034.014, part of the National Roadmap and financed by the Dutch Research Council (NWO).

Data availability

The scripts to produce the figures, raw transport measurement data and TEM data used in this work are available on Zenodo [71].

References

References

  1. [1]D. Jirovec, A. Hofmann, A. Ballabio, P. M. Mutter, G. Tavani, M. Botifoll, A. Crippa, J. Kukucka, O. Sagi, F. Martins, J. Saez-Mollejo, I. Prieto, M. Borovkov, J. Arbiol, D. Chastina, G. Isella, and G. Katsaros, A singlet-triplet hole spin qubit in planar Ge, Nat. Mater. 20, 1106 (2021).DOI
  2. [2]N. W. Hendrickx, D. P. Franke, A. Sammak, G. Scappucci, and M. Veldhorst, Fast two-qubit logic with holes in germanium, Nature 577, 487 (2020).
  3. [3]N. W. Hendrickx, W. I. L. Lawrie, M. Russ, F. van Riggelen, S. L. de Snoo, R. N. Schouten, A. Sammak, G. Scappucci, and M. Veldhorst, A four-qubit germanium quantum processor, Nature 591, 580 (2021).
  4. [4]G. Scappucci, C. Kloefflel, F. A. Zwanenburg, D. Loss, M. Myronov, J.-J. Zhang, S. De Franceschi, G. Katsaros, and M. Veldhorst, The germanium quantum information route, Nat. Rev. Mater. 6, 926 (2021).
  5. [5]Y. Fang, P. Philipppopoulos, D. Culcer, W. A. Coish, and S. Chesi, Recent advances in hole-spin qubits, Mater. Quantum Technol. 3, 012003 (2023).
  6. [6]I. L. Drichko, A. A. Dmitriev, V. A. Malysh, I. Y. Smirnov, H. von Känel, M. Kummer, D. Chastina, and G. Isella, Effective g factor of 2D holes in strained Ge quantum wells, J. Appl. Phys. 123, 165703 (2018).
  7. [7]N. W. Hendrickx, L. Massai, M. Mergenthaler, F. Schupp, S. Paredes, S. W. Bedell, G. Salis, and A. Fuhrer, Sweet-spot operation of a germanium hole spin qubit with highly anisotropic noise sensitivity, Nat. Mater. 23, 920 (2024).
  8. [8]L. A. Terrazos, E. Marcellina, Z. Wang, S. N. Coppersmith, M. Friesen, A. R. Hamilton, X. Hu, B. Koiller, A. L. Saraiva, D. Culcer, and R. B. Capaz, Theory of hole-spin qubits in strained germanium quantum dots, Phys. Rev. B 103, 125201 (2021).DOI
  9. [9]M. Lodari, N. W. Hendrickx, W. I. L. Lawrie, T.-K. Hsiao, L. M. K. Vandersypen, A. Sammak, M. Veldhorst, and G. Scappucci, Low percolation density and charge noise with holes in germanium, Mater. Quantum Technol. 1, 011002 (2021).DOI
  10. [10]M. Myronov, J. Kycia, P. Waldron, W. Jiang, P. Barrios, A. Bogan, P. Coleridge, and S. Studenikin, Holes outperform electrons in group IV semiconductor materials, Small Sci. 3, 2200094 (2023).DOI
  11. [11]L. E. A. Stehouwer, A. Tosato, D. Degli Esposti, D. Costa, M. Veldhorst, A. Sammak, and G. Scappucci, Germanium wafers for strained quantum wells with low disorder, Appl. Phys. Lett. 123, 092101 (2023).DOI
  12. [12]D. Costa, L. E. A. Stehouwer, Y. Huang, S. Martí-Sánchez, D. Degli Esposti, J. Arbiol, and G. Scappucci, Reducing disorder in Ge quantum wells by using thick SiGe barriers, Appl. Phys. Lett. 125, 222104 (2024).
  13. [13]M. Myronov, A. Bogan, and S. Studenikin, Hole mobility in compressively strained germanium on silicon exceeds 7 × 10⁶ cm²V⁻¹s⁻¹, Mater. Today 90, 314 (2025).
  14. [14]F. A. Zwanenburg, A. S. Dzurak, A. Morello, M. Y. Simmons, L. C. L. Hollenberg, G. Klimeck, S. Rogge, S. N. Coppersmith, and M. A. Eriksson, Silicon quantum electronics, Rev. Mod. Phys. 85, 961 (2013).DOI
  15. [15]K. M. Itoh and H. Watanabe, Isotope engineering of silicon and diamond for quantum computing and sensing applications, MRS Commun. 4, 143 (2014).arxiv.org/abs/1410.3922
  16. [16]S. Assali, A. Attiaoui, G. Daligou, P. Daoust, P. Del Vecchio, S. Koelling, L. Luo, N. Rotaru, and O. Moutanabbir, Nuclear spin-depleted, isotopically enriched ⁷⁰Ge/²⁸Si⁷⁰Ge quantum wells, Adv. Mater. 36, 2305703 (2024).DOI
  17. [17]X. Mi, M. Benito, S. Putz, D. M. Zajac, J. M. Taylor, G. Burkard, and J. R. Petta, A coherent spin–photon interface in silicon, Nature 555, 599 (2018).DOI
  18. [18]N. Samkharadze, G. Zheng, N. Kalhor, D. Brousse, A. Sammak, U. C. Mendes, A. Blais, G. Scappucci, and L. M. K. Vandersypen, Strong spin-photon coupling in silicon, Science 359, 1123 (2018).
  19. [19]F. Borjans, X. G. Croot, X. Mi, M. J. Gullans, and J. R. Petta, Resonant microwave-mediated interactions between distant electron spins, Nature 577, 195 (2020).
  20. [20]P. Harvey-Collard, J. Dijkema, G. Zheng, A. Sammak, G. Scappucci, and L. M. K. Vandersypen, Coherent spin-spin coupling mediated by virtual microwave photons, Phys. Rev. X 12, 021026 (2022).
  21. [21]C. X. Yu, S. Zihlmann, J. C. Abadilla-Uriel, V. P. Michal, N. Rambal, H. Niebojewski, T. Bedecarrats, M. Vinet, É. Dumur, M. Filippone, B. Bertrand, S. De Franceschi, Y.-M. Niquet, and R. Maurand, Strong coupling between a photon and a hole spin in silicon, Nat. Nanotechnol. 18, 741 (2023).DOI
  22. [22]J. Dijkema, X. Xue, P. Harvey-Collard, M. Rimbach-Russ, S. L. de Snoo, G. Zheng, A. Sammak, G. Scappucci, and L. M. K. Vandersypen, Cavity-mediated iSWAP oscillations between distant spins, Nat. Phys. 21, 168 (2025).DOI
  23. [23]O. Sagi, A. Crippa, M. Valentini, M. Janik, L. Baghmuyan, G. Fabris, L. Kapoor, F. Hassani, J. Fink, S. Calcaterra, D. Chrastina, G. Isella, and G. Katsaros, A gate tunable transmon qubit in planar Ge, Nat. Commun. 15, 6400 (2024).DOI
  24. [24]A. Nigro, E. Jutzi, F. Oppliger, F. De Palma, C. Olsen, A. Ruiz-Caridad, G. Gadea, P. Scarlino, I. Zardo, and A. Hofmann, Demonstration of microwave resonators and double quantum dots on optimized reverse-graded Ge/SiGe heterostructures, ACS Appl. Electron. Mater. 6, 5094 (2024).DOI
  25. [25]R. M. Lutchyn, E. P. A. M. Bakkers, L. P. Kouwenhoven, P. Krogstrup, C. M. Marcus, and Y. Oreg, Majorana zero modes in superconductor–semiconductor heterostructures, Nat. Rev. Mater. 3, 52 (2018).DOI
  26. [26]E. Prada, P. San-Jose, M. W. A. de Moor, A. Geresdi, E. J. H. Lee, J. Klinovaja, D. Loss, J. Nygaard, R. Aguado, and L. P. Kouwenhoven, From Andreev to Majorana bound states in hybrid superconductor–semiconductor nanowires, Nat. Rev. Phys. 2, 575 (2020).
  27. [27]M. Hays, V. Fatemi, D. Bouman, J. Cerrillo, S. Diamond, K. Serniak, T. Connolly, P. Krogstrup, J. Nygård, A. Levy Yeyati, A. Geresdi, and M. H. Devoret, Coherent manipulation of an Andreev spin qubit, Science 373, 430 (2021).DOI
  28. [28]M. Pita-Vidal, A. Bargerbos, R. Žitko, L. J. Splitthoff, L. Grünhaupt, J. J. Wesdorp, Y. Liu, L. P. Kouwenhoven, R. Aguado, B. van Heck, A. Kou, and C. K. Andersen, Direct manipulation of a superconducting spin qubit strongly coupled to a transmon qubit, Nat. Phys. 19, 1110 (2023).
  29. [29]S. Hoffman and C. Tahan, Resolving Andreev spin qubits in germanium-based Josephson junctions (2025), arXiv:2506.13988.arxiv.org/abs/2506.13988
  30. [30]T. Dvir, G. Wang, N. van Loo, C.-X. Liu, G. P. Mazur, A. Bordin, S. L. D. ten Haaf, J.-Y. Wang, D. van Driel, F. Zatelli, X. Li, F. K. Malinowski, S. Gazibegovic, G. Badawy, E. P. A. M. Bakkers, M. Wimmer, and L. P. Kouwenhoven, Realization of a minimal Kitaev chain in coupled quantum dots, Nature 614, 445 (2023).
  31. [31]K. Laubscher, J. D. Sau, and S. Das Sarma, Majorana zero modes in gate-defined germanium hole nanowires, Phys. Rev. B 109, 035433 (2024).DOI
  32. [32]K. Laubscher, J. D. Sau, and S. Das Sarma, Germanium-based hybrid semiconductor-superconductor topological quantum computing platforms: Disorder effects, Phys. Rev. B 110, 155431 (2024).DOI
  33. [33]L. Lakic, W. I. L. Lawrie, D. van Driel, L. E. A. Stehouwer, Y. Su, M. Veldhorst, G. Scappucci, F. K. Kuemeth, and A. Chatterjee, A quantum dot in germanium proximitized by a superconductor, Nat. Mater. 24, 552 (2025).DOI
  34. [34]N. van Loo, G. P. Mazur, T. Dvir, G. Wang, R. C. Dekker, J.-Y. Wang, M. Lemang, C. Slgijog, A. Bordin, D. van Driel, G. Badawy, S. Gazibegovic, E. P. A. M. Bakkers, and L. P. Kouwenhoven, Electrostatic control of the proximity effect in the bulk of semiconductor-superconductor hybrids, Nat. Commun. 14, 3325 (2023).
  35. [35]F. Vigneau, R. Mizokuchi, D. C. Zanuz, X. Huang, S. Tan, R. Maurand, S. Frolov, A. Sammak, G. Scappucci, F. Lefloch, and S. De Franceschi, Germanium quantum-well Josephson field-effect transistors and interferometers, Nano Lett. 19, 1023 (2019).DOI
  36. [36]J. Ridderbos, M. Brauns, F. K. de Vries, J. Shen, A. Li, S. Kölling, M. A. Verheijen, A. Brinkman, W. G. van der Wiel, E. P. A. M. Bakkers, and F. A. Zwanenburg, Hard superconducting gap and diffusion-induced superconductors in Ge–Si nanowires, Nano Lett. 20, 122 (2020).
  37. [37]K. Aggarwal, A. Hofmann, D. Jirovec, I. Prieto, A. Sammak, M. Botifoll, S. Martí-Sánchez, M. Veldhorst, J. Arbiol, G. Scappucci, J. Danon, and G. Katsaros, Enhancement of proximity-induced superconductivity in a planar Ge hole gas, Phys. Rev. Res. 3, L022005 (2021).
  38. [38]A. Tosato, V. Levajac, J.-Y. Wang, C. J. Boor, F. Borsoi, M. Botifoll, C. N. Borja, S. Martí-Sánchez, J. Arbiol, A. Sammak, M. Veldhorst, and G. Scappucci, Hard superconducting gap in germanium, Commun. Mater. 4, 23 (2023).DOI
  39. [39]M. Hinderling, S. C. ten Kate, M. Coraiola, D. Z. Haxell, M. Stiefel, M. Mergenthaler, S. Paredes, S. W. Bedell, D. Sabonis, and F. Nichele, Direct microwave spectroscopy of Andreev bound states in planar Ge Josephson junctions, PRX Quantum 5, 030357 (2024).DOI
  40. [40]P. Viswanathan, E. Lemmens Sjöstrand, D. Costa, M. Fischer, A. Ahmed, S. Martí-Sánchez, L. E. A. Stehouwer, J. Arbiol, A. Chatterjee, G. Scappucci, and K. L. Hudson, Buried germanium quantum well proximitised by magnetic field-resilient superconducting platinum iridium germanosilicide (2026), arXiv:2608.09611.arxiv.org/abs/2608.09611
  41. [41]H. Li, Z. Shang, M. P. Lilly, M. Myronov, and L. P. Rokhinson, Characterization of superconducting germanide and germanosilicide films of Pd, Pt, Rh and Ir formed by solid-phase epitaxy, APL Mater. 14, 031101 (2026), arXiv:2509.14173.arxiv.org/abs/2509.14173
  42. [42]B. T. Matthias, T. H. Geballe, and V. B. Compton, Superconductivity, Rev. Mod. Phys. 35, 1 (1963).pubmed.ncbi.nlm.nih.gov/17799958
  43. [43]A. K. Ghosh and D. H. Douglass, Superconductivity in transition metal-germanium systems, J. Low Temp. Phys. 27, 487 (1977).DOI
  44. [44]G. F. Hardy and J. K. Hulm, The superconductivity of some transition metal compounds, Phys. Rev. 93, 1004 (1954).
  45. [45]J. R. Gavaler, Superconductivity in Nb–Ge films above 22 K, Appl. Phys. Lett. 23, 480 (1973).DOI
  46. [46]L. R. Testardi, Structural instability and superconductivity in A-15 compounds, Rev. Mod. Phys. 47, 637 (1975).DOI
  47. [47]G. R. Stewart, Superconductivity in the A15 structure, Physica C 514, 28 (2015).DOI
  48. [48]D. A. Cardwell, D. C. Larbalestier, and A. I. Braginski, Handbook of Superconductivity: Processing and Cryogenics, Volume Two, 2nd ed. (CRC Press, 2022).DOI
  49. [49]N. Terada, Y. Hoshi, M. Naoe, and S. Yamanaka, Preparation of high T_c Nb-(Ge,Si) thin films at low substrate temperature, in Advances in Cryogenic Engineering Materials: Volume 30, edited by A. F. Clark and R. P. Reed (Springer US, 1984) pp. 559–570.DOI
  50. [50]P. J. Strohbheen, T. Banerjee, A. M. Brook, I. Levy, W. L. Sarney, J. van Dijk, H. Orth, M. Mikalsen, V. Fatemi, and J. Shabani, Molecular beam epitaxy growth of superconducting tantalum germanide, Appl. Phys. Lett. 124, 092102 (2024).DOI
  51. [51]C. Kopas, S. Zhang, J. Gonzales, D. R. Queen, B. Wagner, R. W. Carpenter, and N. Newman, Ge thin-films with tantalum diffusion-barriers for use in Nb-based superconductor technology (2021), arXiv:2104.12580 [cond-mat.mtrl-sci].arxiv.org/abs/2104.12580
  52. [52]E. Ahmed, K. Kobayashi, N. Arakawa, Y. Okubo, A. Sakai, S. Nakatsujii, A. Iyo, and T. Ebihara, Superconducting properties of the A15 structure compound V3Ge, Physica C 602, 1354110 (2022).DOI
  53. [53]L. A. Pendrys and D. H. Douglass, Studies of superconducting A-15 vanadium-based alloys, J. Low Temp. Phys. 23, 367 (1976).DOI
  54. [54]T. B. Massalski, H. Okamoto, P. R. Subramanian, and L. Kacprzak, eds., Binary Alloy Phase Diagrams, 2nd ed. (ASM International, Materials Park, OH, 1990).
  55. [55]F. M. d’Heurle and P. Gas, Kinetics of formation of silicides: A review, J. Mater. Res. 1, 205 (1986).
  56. [56]M.-A. Nicolet and S. S. Lau, Formation and characterization of transition-metal silicides, in VLSI Electronics: Microstructure Science, Vol. 6, edited by N. G. Einspruch and G. B. Larrabee (Academic Press, New York, 1983) pp. 329–464.DOI
  57. [57]G. Gottlieb, J. C. Lasjaunias, J. L. Tholence, O. Laborde, O. Thomas, and R. Madar, Superconductivity in TaSi2 single crystals, Phys. Rev. B 45, 4803 (1992).pubmed.ncbi.nlm.nih.gov/10002118
  58. [58]Z. Li, W.-W. Yang, J.-J. Feng, B.-B. Ruan, Y.-Z. Zhou, T. Sun, H.-W. Shi, L. Ju, S.-F. Wang, T. Han, J.-P. Song, X.-R. Hu, Z.-K. Zhou, M.-S. Long, X.-Y. Hou, Q.-N. Xu, Z.-A. Ren, Q.-G. Mu, and L. Shan, Type-I superconductivity in the Weyl semimetal TaGe2 with chiral structure, Phys. Rev. B 110, 174506 (2024).DOI
  59. [59]J. J. Hanak, J. I. Gittleman, J. P. Pellicane, and S. Bozowski, Radio-frequency-sputtered films of β-tungsten structure compounds, J. Appl. Phys. 41, 4958 (1970).DOI
  60. [60]E. Emmanouilidou, S. Mardanya, J. Xing, P. V. S. Reddy, A. Agarwal, T.-R. Chang, and N. Ni, Fermiology and type-I superconductivity in the chiral superconduc- tor NbGe₂ with Kramers-Weyl fermions, Phys. Rev. B 102, 235144 (2020).
  61. [61]J. C. Lasjaunias, O. Laborde, U. Gottlieb, R. Madar, and O. Thomas, Low temperature specific heat of VSi₂, NbSi₂, and TaSi₂, J. Low Temp. Phys. 92, 335 (1993).DOI
  62. [62]V. M. Kuzʼmenko, V. G. Lazarev, V. I. Melʼnikov, and A. I. Sudovtsov, Superconductivity of amorphous vanadium, Sov. Phys. JETP 40, 396 (1975).
  63. [63]M. H. Read and C. Altman, A new structure in tantalum thin films, Appl. Phys. Lett. 7, 51 (1965).DOI
  64. [64]B. W. Roberts, Survey of superconductive materials and critical evaluation of selected properties, J. Phys. Chem. Ref. Data 5, 581 (1976).DOI
  65. [65]B. Langa, D. Sakpata, I. Lainez, R. Haight, B. Srijanto, L. Feldman, H. Hijazi, X. Zhu, L. Hu, M. Kim, and K. Sardashti, Solid-state reactions at niobium–germanium interfaces in hybrid quantum electronics, AIP Adv. 14, 095311 (2024)
  66. [66]H. Kräutle, M.-A. Nicolet, and J. W. Mayer, Kinetics of silicide formation by thin films of V on Si and SiO₂ substrates, J. Appl. Phys. 45, 3304 (1974).DOI
  67. [67]S. Gaudet, C. Detavernier, A. J. Kellock, P. Desjardins, and C. Lavoie, Thin film reaction of transition metals with germanium, J. Vac. Sci. Technol. A 24, 474 (2006).DOI
  68. [68]P. Kringhøj, A. Nylandsted Larsen, and S. Y. Shiryaev, Diffusion of Sb in strained and relaxed Si and SiGe, Phys. Rev. Lett. 76, 3372 (1996).pubmed.ncbi.nlm.nih.gov/10060950
  69. [69]H. Bracht, S. Schneider, and R. Kube, Diffusion and doping issues in germanium, Microelectron. Eng. 88, 452 (2011).DOI
  70. [70]A. Chroneos and H. Bracht, Diffusion of n-type dopants in germanium, Appl. Phys. Rev. 1, 011301 (2014).DOI
  71. [71]S. R. Roelofs, Data, code and figures for ”solid-state reaction synthesis of superconducting refractory-metal germanides for ge/sige quantum devices” (2026).DOI

Paper details

vlnc:W_djewb7en
Version
v1 · latest
Published
Fresh — added to the Hub within the last 30 days
Added
License
CC BY 4.0
Usage

views · 10

downloads · 0

A view counts once per browser per day, a download each time the PDF is requested.

  • Physics and Astronomy

Not recorded for this paper

View statement
Looking for related papers…
Contents