Journal of the Physical Society of Japan Nodal Superconductivity of UTe2 Probed by Field-Angle-Resolved Specific Heat on a Crystal with Tc = 2.1 K Kaito Totsuka1,2 , Yohei Kono2 , Yusei Shimizu3,4 , Ai Nakamura4 , Atsushi Miyake4 , Dai Aoki4 , Yasumasa Tsutsumi5 , Kazushige Machida6 , and Shunichiro Kittaka1,2 * 1 arXiv:2601.08201v2 [cond-mat.supr-con] 20 Feb 2026 2 Department of Basic Science, The University of Tokyo, Meguro, Tokyo 153-8902, Japan Department of Physics, Faculty of Science and Engineering, Chuo University, Kasuga, Bunkyo-ku, Tokyo 112-8551, Japan 3 Institute for Solid State Physics, The University of Tokyo, Kashiwa, Chiba 277-8581, Japan 4 IMR, Tohoku University, Oarai, Ibaraki 311-1313, Japan 5 Department of Civil and Environmental Engineering, Faculty of Engineering, Hiroshima Institute of Technology, Hiroshima 731-5193, Japan 6 Department of Physics, Ritsumeikan University, Kusatsu, Shiga 525-8577, Japan Field-angle-resolved specific-heat measurements were performed on a clean single crystal of a spin-triplet superconductor UTe2 with T c = 2.1 K and a low residual electronic specific heat. At low temperatures, the specific heat exhibits a linear dependence on the magnetic field when the field is applied precisely along the b axis, in stark contrast to its rapid increase at low fields for other orientations. This pronounced anisotropy suggests the presence of nodal quasiparticle excitations with the Fermi velocity predominantly aligned along the b axis. Considering the characteristic field-angle dependences of both the specific heat and the upper critical field, these observations are broadly compatible with theoretical models that assume a superconducting gap structure featuring either point nodes consistent with B2u symmetry, allowed in the infinitely strong spin-orbit coupling scheme, or line nodes confined to flat regions of the quasi-twodimensional Fermi surface, consistent with 3 B3u symmetry in the finite spin-orbit classification scheme. These results yield crucial hints for resolving the pairing symmetry of UTe2 , paving the way for a deeper understanding of its spintriplet superconductivity. 1. Introduction ing gap structure, remains controversial. Various experimental techniques, including thermal conductivity κ,13, 14) magnetic penetration depth,15) quasiparticle interference,16) and specific heat,17) have yielded seemingly conflicting evidence for both fully-gapped and nodal superconductivity. In addition, while magnetic penetration depth measurements suggest a multi-component order parameter,15) Kerr-effect and ultrasound studies point toward a single-component scenario in the orbital part of the pairing function.18, 19) Furthermore, the geometry of the Fermi surface (FS), whether predominantly two-dimensional or three-dimensional, adds another layer of complexity to the debate.20–24) To address these unresolved issues, particularly the nodal structure of the superconducting gap, we perform lowtemperature specific-heat measurements on a clean single crystal of UTe2 , with precise control of the magnetic-field orientation. Our results provide thermodynamic evidence for low-energy excitations of nodal quasiparticles whose Fermi velocity is oriented along the b axis. Spin-triplet superconductivity, distinguished by its rich internal degrees of freedom and potential for hosting unconventional quantum states, stands at the forefront of contemporary condensed matter physics. Despite decades of theoretical predictions and experimental efforts, the identification of candidate materials has remained a formidable challenge, leaving the fundamental nature of this exotic pairing largely unresolved. The discovery of superconductivity in UTe2 marked a turning point in this quest.1–3) With its extraordinary superconducting properties, including a high upper critical field far exceeding the Pauli limit, field-reentrant superconductivity above 40 T,4) and multiple superconducting phases induced by magnetic field and pressure,5) UTe2 has rapidly emerged as a premier platform for investigating spin-triplet pairing. UTe2 crystallizes in the orthorhombic space group Immm with the a axis being the magnetization easy axis. Shortly after its discovery, the superconducting transition temperature T c was reported to be 1.6 K. However, recent advances in crystal growth techniques have led to the synthesis of nextgeneration single crystals, in which T c is significantly enhanced to 2.1 K.6, 7) These high-quality crystals, characterized by an exceptionally low residual electronic specific-heat coefficient in the superconducting state, enable precise investigations of the intrinsic properties of UTe2 without the complications arising from impurity and inhomogeneity. While the spin component has been strongly suggested to be of triplet character, based on comprehensive NMR studies,8–12) the orbital part of the pairing function, namely, the superconduct- 2. Methods High-quality single crystals were grown using the molten salt flux liquid transport method.7) For this study, a thin, rectangular single crystal with a mass of 11.68 mg, elongated along the a axis, was selected. The surface of the sample parallel to the ab plane was securely attached to the calorimeter stage using GE varnish. Subsequently, the calorimeter was mounted on a dilution refrigerator (Oxford, Kelvinox AST Minisorb) such that the c axis of the sample was aligned with the vertical z direction. This refrigerator was inserted into a vector magnet system,25) which can generate magnetic fields * kittaka@g.ecc.u-tokyo.ac.jp 1 J. Phys. Soc. Jpn. 3. (a) 400 100 2T 1T 0T 200 B || c UTe2 0 0 1 T (K) 2 C / T (mJ / mol K2) C / T (mJ / mol K2) up to 7 T (3 T) along the horizontal x (z) direction. The refrigerator can be rotated around the z axis using a stepper motor. Using this system, we achieved three-dimensional control of the magnetic-field orientation with an angular resolution better than 0.1◦ . The heat capacity was measured using the quasiadiabatic heat-pulse method. The addenda heat capacity was measured separately and exhibited a Schottky-type anomaly in its field dependence at low temperatures.26) This addenda contribution has been subtracted from all data presented below. c 50 (b) 0.3 K a b 0 0 50 1 2 B0.5 (T0.5) c a b 0 0 1 2 3 B (T) 4 5 Results and Discussion Fig. 1. (Color online) (a) Temperature dependence of C/T under magnetic fields applied along the c axis. (b) Magnetic-field dependence of C/T at 0.3 K for √ each field orientation. The inset shows the same data plotted as a function of B. 3.1 Temperature dependence of the specific heat Figure 1(a) shows the temperature dependence of the specific heat divided by temperature, C/T , where the phonon contribution has been subtracted. In zero field, a sharp specific-heat jump, ∆C/(γT c ) = 2.5, is observed at T c = 2.05 K with a transition width of about 0.06 K (onset at 2.1 K), where γ = 125 mJ mol−1 K−2 is the electronic specific heat coefficient in the normal state. These results, together with the small value of C/(γT ) = 0.06 at 0.1T c , demonstrate the high quality of the present sample. As exemplified by the data for B ∥ c in Fig. 1(a), the temperature dependence of C/T under magnetic fields for each orientation generally resembles the behavior reported in Ref. 17, which investigated a sample with T c = 2.1 K and C/(γT ) ∼ 0.12 at 0.1T c in zero field. In particular, our data reproduce the anomalous T 2 dependence of C/T for B ∥ a above 1 T, as reported in Supplementary Fig. S4 of Ref. 17, whose contribution becomes more pronounced with increasing temperature. In addition, an upturn in the specific heat was observed below 0.15 K, similar to that previously reported for low-T c samples,27) which is unlikely to originate from the nuclear contribution. The origins of both the T 2 behavior and the low-temperature upturn remain unclear. To minimize these contributions, field-angle-resolved specificheat measurements in this study were primarily performed at 0.3 K (∼ 0.15T c ). At this temperature, the influence of anomalous components is reduced, enabling the detection of low-energy quasiparticle excitations around nodes, as demonstrated in previous studies on various unconventional superconductors.28–31) energy quasiparticles are excited due to the Doppler energy shift given by ∆E = me vF · vs , where me is the electron mass, vF is the Fermi velocity, and vs is the local superfluid velocity circulating around vortex cores, which is perpendicular to the√field direction. This so-called Volovik effect33) leads to a B-like behavior in the zero-energy quasiparticle density of states (ZDOS), N(E = 0), which can be detected through low-temperature specific-heat measurements. However, when vF at the nodes is uniquely oriented and the magnetic field is applied along this nodal vF direction, i.e., vs ⊥ vF , low-energy quasiparticle excitations are suppressed due to ∆E = 0. In this case, a linear dependence N(E = 0) ∝ B is expected, as it reflects the number of vortex cores. Based on the simple framework described above, the results shown in Fig. 1(b) provide firm thermodynamic evidence that nodes exist on the FS, where vF points along the b axis, and are absent on regions with vF pointing in other directions. 3.3 Field-angle dependence of the specific heat In order to further investigate the gap structure, we have investigated the field-angle dependence of the specific heat. Indeed, the field-angle dependence of N(E = 0) and/or C for various types of nodal superconductors has been extensively studied from a theoretical perspective.32, 34–37) It should be noted that the FS geometry affects the specific-heat oscillations associated with gap anisotropy, including characteristic fields and temperatures where the oscillation sign reverses.35) For uniaxially symmetric point-node or line node gap structures on a three-dimensional FS, the maxima of the field-angle-dependent ZDOS N(θ) shift from the antinodal to the nodal direction with increasing magnetic field, accompanied by a reversal of the oscillation pattern between low and high fields.37) Therefore, to reliably determine the nodal position and structure, it is crucial to analyze the field-angle dependence of the specific heat over a range of magnetic fields, taking into account the possible reversal of the oscillation pattern. Figures 2(a)-2(c) show the specific heat measured at 0.3 K under magnetic fields rotated within the ac, ab, and bc planes, respectively. Here, the field angle θ (ϕ) denotes the polar angle (azimuthal angle in the ab plane) measured from the c (a) axis. In each rotational plane, the specific heat exhibits a char- 3.2 Magnetic field dependence of the specific heat Figure 1(b) shows the magnetic-field dependence of the specific heat for fields applied along each axis at 0.3 K. When the magnetic field is applied along the √a or c axis, a rapid increase in C(B), reminiscent of the B (B0.64 ) behavior expected for line-node (point-node) superconductors, is observed in the low-field region,32) as confirmed in the inset of √ Fig. 1(b). Although Ref. 17 reported a B-like dependence of the specific heat for all magnetic-field directions, our measurements, with fine tuning of the field orientation and small field increments, revealed that C(B) is nearly proportional to B along the b axis. The previous results obtained using a lowT c sample with C/(γT ) ∼ 0.6 at 0.1T c in zero field27) qualitatively differ from the present findings. This discrepancy is likely attributable to impurity effects in the earlier sample, which may have masked the signatures of low-energy quasiparticle excitations. When the superconducting gap possesses nodes, low2 J. Phys. Soc. Jpn. a 0.3 K C / T (mJ / mol K2) 60 c a 3 a 5 c 1 (a) 0 0 T (b) 90 180 θ (deg) 0 0 T (c) 90 180 (deg) 0 b 120 0.5 0.5 0.2 c c c 140 1 0.5 20 140 2 2 1 c 3 3 40 b 0.3 K 4.5 4 2 0 b C / T (mJ / mol K2) c 0T 90 180 θ (deg) 100 40 c 2.3 K 2 T (b) 0 a 0.9 K 60 0 120 0.5 T 80 20 Fig. 2. (Color online) Field-angle dependence of C/T at 0.3 K under various magnetic fields rotated within the (a) ac, (b) ab, and (c) bc planes. The numbers labeling each set of data indicate the magnetic-field strength in tesla. Open symbols represent data points mirrored with respect to the crystal symmetry axes. 1.2 K a 140 120 0.6 K (a) 0.3 K 0 90 180 θ (deg) 140 120 90 180 θ (deg) b a 2.3 K 2 T (c) 0 90 180 (deg) c b c 2.3 K 2 T (d) 0 90 180 θ (deg) Fig. 3. (Color online) (a) Field-angle dependence of C/T in the superconducting state at various temperatures under a magnetic field of 0.5 T rotated within the bc plane. (b)-(d) C/T in the normal state at 2.3 K under a magnetic field of 2 T rotated within the (b) ac, (c) ab, and (d) bc planes. Open symbols indicate data points mirrored with respect to the crystal symmetry axes. acteristic, field-dependent oscillation pattern, as described in detail below. 3.3.1 ac-plane field rotation We first focus on the oscillation pattern of the specific heat C(θ) obtained for field rotation within the ac plane at ϕ = 0◦ , as shown in Fig. 2(a). At low magnetic fields, C(θ) exhibits a minimum at θ = 90◦ (B ∥ a), which remains unchanged at higher fields. Above 0.75 T, a dip develops at θ = 0◦ (B ∥ c), accompanied by a local maximum around 30◦ ≲ θ ≲ 45◦ . Based on the C(B) data in Fig. 1(b), no nodes can exist within the ac plane. This suggests that nodal quasiparticles do not significantly contribute to the anisotropy in C(θ) at ϕ = 0◦ . Therefore, the observed oscillation in Fig. 2(a) is most reasonably attributed to the anisotropy of the upper critical field Bc2 , as C(B) generally scales with B/Bc2 . In this material, the anisotropy of Bc2 within the ac plane exhibits a pronounced temperature and field dependence:38) ∥c near T c and at low fields, B∥a c2 > Bc2 , whereas at low tem∥a peratures and high fields, the relation reverses to B∥c c2 > Bc2 . This behavior suggests a possible inversion of superconducting anisotropy with increasing field, although its microscopic origin remains unresolved. The observed θ dependence of the specific heat within the ac plane can thus be interpreted as part of this crossover process. Indeed, Ref. 17 reports a reversal of the specific heat anisotropy around 6 T. Therefore, the crossover in the oscillation pattern of C(θ) for B ∥ ac with increasing field strength likely reflects a field-induced change in the superconducting anisotropy, rather than a change driven by nodal quasiparticle excitations. Across the present field range, C(ϕ) at θ = 90◦ consistently exhibits a minimum along the b axis. Above 3 T, a dip develops around the a axis (ϕ = 0◦ ); the maximum shifts from the a toward the b axis, accompanied by a shoulder or peak anomaly at intermediate field angles. Unlike Bc2 (θ) in the ac plane, Bc2 (ϕ) in the ab plane does not show a prominent minimum at intermediate field angles.38) Therefore, the observed oscillations in C(ϕ) cannot be explained solely by the anisotropy of Bc2 . Theoretical models assuming a spherical FS and the presence of nodes with vF ∥ b predict a shoulder or peak anomaly in the angular dependence of N(E = 0) within the ab plane, appearing at an intermediate field angle.27, 31, 37) The consistency between experimental results and theoretical predictions supports the presence of nodes along vF ∥ b, as also indicated by the C(B) data. 3.3.3 bc-plane field rotation In contrast to the non-monotonic field-angle dependence of C(ϕ) for B ∥ ab, C(θ) at ϕ = 90◦ (B ∥ bc) [Fig. 2(c)] consistently exhibits a | cos θ|-like behavior with a minimum along the b axis, without showing a shoulder or peak anomaly. Since the anisotropy of the coherence length at low fields is relatively small, this sharp-dip structure around B ∥ b cannot be attributed to Bc2 anisotropy, but rather suggests lowenergy quasiparticle excitations associated with point nodes or line nodes.37, 39) With increasing temperature, the sharp-dip feature in the | cos θ|-like behavior is gradually suppressed, as shown in Fig. 3(a). This suppression may be attributed to thermal quasiparticle excitations around nodes where vF ∥ b, which becomes active upon warming even when B ∥ b. In the normal state, as shown in Figs. 3(b)-3(d), no prominent field-angle dependence is observed in C(ϕ, θ), in contrast to a previous report using a low-T c sample.27) This demonstrates that the low-temperature oscillation in C(ϕ, θ) is an intrinsic feature of the superconducting state. 3.3.2 ab-plane field rotation We next turn to the results of C(ϕ) for B ∥ ab, shown in Fig. 2(b). Assuming that nodes exist only along the b axis, quasiparticle excitations are nearly isotropic when the field is rotated within the ac plane. In contrast, for rotations within planes that include the b axis, the field direction relative to the node strongly affects quasiparticle excitations, making the influence of the gap structure more pronounced than in the ac plane. 3 J. Phys. Soc. Jpn. _ (a) kb + α (c) ka β b 5 0 -5 c -5 0 N(E=0)/N0 5 0 -5 c a -5 0 a 0.4 5 b (e) 5 b 00 0 0.1 0.2 B 0.3 B || c a 0 -5 0.5 0.2 0 c (d) 5 B || b (b) 0.6 √ shift ∆E ∝ vF · vs , giving rise to the Volovik B effect in γ(B) for B ∥ b. However, quantum oscillation experiments 23, 24, 42, 43) and band calculations44) have revealed exceptionally flat FS regions on the b face of the β sheet and a face of the α sheet. The flatness of the b face has also been confirmed by angular-dependent magnetoresistance oscillation measurements, specifically through the observation of the so-called Yamaji oscillations.45) Eaton et al.24) have constructed an analytical FS model (the Eaton model) consistent with these observations. In our calculations, we adopt the Eaton model24) to examine the quasiparticle excitations and to narrow down the possible pairing symmetries realized in UTe2 . To best describe the experimental data discussed above, we assume a gap function ∆(k) = ∆0 sin(aka /2) where a is the lattice constant along the a axis. This gap structure has line nodes on the b face of the β sheet, as indicated by circles in Fig. 4(a). In contrast, the α sheet remains fully gapped, although the gap function exhibits a sign change within a unit cell, as illustrated by the different background colors labeled with ± in Fig. 4(a). Figure 4(b) presents the calculated field dependence of the ZDOS, N(E = 0)/N0 , or equivalently γ(B), for the three magnetic-field orientations. Here, N0 denotes the density of states in the normal √ state. It is evident that γ(B) for B ∥ c and B ∥ a exhibit a B dependence, characteristic of the Volovik effect associated with nodal quasiparticle excitations, as high√ lighted in the inset where the data are plotted against B. In contrast, γ(B) for B ∥ b increases nearly linearly with B, indicating that no Doppler shift occurs because vF (kc ) at the nodal positions is perpendicular to vs along the kc axis. Therefore, the observed strict linear behavior C(B) ∝ B for B ∥ b, as shown in Fig. 1(b), implies that the FS hosting the line nodes must be flat, with no dispersion of vF (kc ) along the kc direction. The particular requirement is fulfilled by only the b face of the β sheet. In such a case, the gap amplitude does not depend on the field angle within the bc plane, where the magnetic field is rotated [see Fig. 2(c)], and no field-angle dependence of the specific heat arising from gap anisotropy would be expected. Although the a face of the α sheet also satisfies this requirement, √ as seen in Fig. 4(a), the experimental fact that C(B) ∝ B for B ∥ a excludes the possibility of line nodes along the a axis. We also note that the b face of the α sheet has strong warping along the kc direction, making that the α sheet is not eligible √ for possessing line nodes. Such warping would lead to a B dependence in γ(B) for B ∥ b, which contradicts the observed linear behavior. As shown in Fig. 4(c) for B ∥ a and Fig. 4(e) for B ∥ c, the zero-energy quasiparticle spectral weight well extends toward the nodal directions along the kb axis. In contrast, for B ∥ b [Fig. 4(d)], the spectral weight is confined near the vortex core at the center, still leaking √ out toward the offaxis directions. This results in a small B behavior barely seen in the lowest field region of Fig. 4(b), although such behavior is not actually observed in Fig. 1(b). The open angle of ZDOS is determined by the vF anisotropy, namely tan−1 (ṽcF /ṽaF ) = tan−1 (0.25/0.70) ∼ 20◦ , where ṽcF denotes the root-mean-square average vcF over the FS evaluated in the Eaton model. We notice that in recent STM experiments46–48) performed for B ∥ (01̄1) on the (011) surface, which is tilted only by ∼24◦ from the b axis, have revealed a similar ZDOS feature elongated along the a direction and centered at the B || a 2 B0.5 0.6 0.4 0.5 1 -5 0 5 N(E=0) N0 0 Fig. 4. (Color online) (a) Top view of the 3D Fermi surfaces projected onto the ka -kb plane. The thickness of the FS lines is roughly proportional to the warping along the kc axis. The red circles indicate the positions of the line nodes for 3 B3u . Point nodes are located at both circles and squares for B2u . The sign of the order parameter is alternate along the ka direction. The original figure is drawn by Eaton et al.24) (b) Calculated results of the ZDOS, N(E = 0)/N0 or γ(B), as a function of B for three √ field orientations. The inset shows the same data plotted as a function of B. (c)-(e) The spatial profiles of the ZDOS centered at the vortex core in a unit cell for three field directions at B = 0.04. Here, B is expressed in the Eilenberger unit37) and T = 0.2T c . 3.4 Possible superconducting gap structures 3.4.1 Point-node scenario The absence of a shoulder or peak anomaly in C(θ) for B ∥ bc is apparently incompatible with the theoretical prediction37) which assumes that quasiparticle excitations are dominated by point nodes located along vF ∥ b on a threedimensional FS. This discrepancy may arise from the geometry of the FS. If the dominant superconducting bands in UTe2 are quasi-two-dimensional FSs extended along the c axis, there would be many regions where vF ∥ b, whereas regions with vF ∥ c would be absent. It should be noted that a non-monotonic oscillation pattern, often accompanied by a shoulder or peak anomaly, can occur when the N(E = 0) or γ(B) ≡ C(B)/T |T →0 curves for B ∥ b and B ∥ c approach each other and intersect. As seen in Fig. 1(b), the C(B) curves for B ∥ b and B ∥ c show no sign of crossing, even at higher fields up to Bc2∥c ∼ 14 T.17) 3.4.2 Line-node scenario We now examine the alternative possibility of the line node scenario relative to the point node one by backing up the microscopic quasi-classical Eilenberger theory40, 41) with a realistic FS model, consisting of the two cylindrical FSs24) as shown in Fig. 4(a). The FS information is critical in performing a realistic construction of the quasiparticle excitations in a vortex state to understand the angle-resolved specific heat data shown above. It is not obvious to explain the linear field dependence of C(B) ∝ B for B ∥ b because, when the magnetic field is applied along the nodal direction, the line nodes on the three dimensional FS necessarily have a component of vF parallel to vs . This leads to the Doppler 4 J. Phys. Soc. Jpn. vortex core. The above calculated results uniquely identify the position of the line nodes described by ∆(k) = ∆0 sin(aka /2). Namely, the line nodes must reside on the b face of the β sheet. As shown in Fig. 4(a), the gap function yields that (1) the line nodes are located on the β sheet, and (2) the α sheet is fully gapped with a sign change. This uniqueness arises from the particular FS structure, which has been determined through both experimental and theoretical investigations.23, 24, 42–45) Away from the kb axis, the β sheet begins to warp. Any √ nodal structures deviating from the b axis would exhibit a B behavior for B ∥ b, which is incompatible with the strictly linear B dependence observed in C(B). nodes emerge only in regions where vF is nearly parallel to the b axis. For example, previous penetration depth study supports the chiral B3u + iAu state in the zero-field limit, which possesses point nodes near the ky and kz axes.15) This nodal structure is unlikely based on the present results. However, the application of a magnetic field can modify the chiral order parameter, potentially altering the positions of nodes under finite fields. As discussed by Ishihara et al.,15) this possibility may resolve the apparent contradiction between our specific heat data and previous penetration depth study. By contrast, recent thermal-conductivity measurements13) report a vanishingly small value of κ/T and a lack of field dependence at low fields in the zero-temperature limit, which apparently contradict nodal superconductivity. These results have been interpreted as evidence for a fully gapped superconducting state with Au symmetry. One possible explanation for this discrepancy lies in thermal conductivity is primarily sensitive to light-mass quasiparticles, whereas specific heat captures contributions from the entire FSs, including heavymass quasiparticles. Moreover, thermal-conductivity measurements at low temperatures may be affected by electronphonon decoupling, potentially leading to an underestimation of the electronic contribution.56) Further studies are required to resolve the apparent discrepancy between the b-axis nodal scenario indicated by the present specific-heat measurements and the fully gapped state suggested by thermal conductivity data. 3.5 Possible gap symmetries Let us discuss possible gap symmetries of UTe2 . The present specific-heat results suggest the existence of nodes along the direction of vF ∥ b. If the gap nodes are point-like, the only compatible single-component gap symmetry within the D2h point group, assuming an infinitely strong spin-orbit coupling (SOC) scheme, is the B2u representation.49) Both circles and squares in Fig. 4(a) indicate the positions of the point nodes. This symmetry is consistent with previous reports based on ultrasound and thermal-conductivity measurements.14, 19) Alternatively, a line-node gap symmetry is possible if one assumes a flat FS. This type of line-node gap has not been widely discussed, primarily because it appears to violate the Blount theorem.50) However, the theorem is strictly applicable only to classification schemes for triplet pairing in the limit of infinitely strong SOC. According to the Blount theorem, symmetry-protected line nodes are forbidden under strong SOC, where spin and orbital degrees of freedom are tightly coupled, except for non-symmorphic crystals,51) which is not the case for UTe2 . Conceptually, this apparent contradiction can be resolved by reconsidering the classification scheme: from the strong SOC limit, described by spin+orbital D2h × U(1), to a finite SOC framework, described by SO(3)spin × Dorbital × U(1).49, 52, 53) In the strong SOC scheme, 2h the Cooper pair moment is locked to the crystal lattice and cannot reorient under an applied magnetic field. In contrast, the finite SOC scheme allows for d-vector rotation, consistent with experimental observations at ∼ 3 T (14 T) for B ∥ c (B ∥ b).8–12) Importantly, the finite SOC scheme generically permits the existence of line nodes.49, 52) The identified pairing function with line nodes along the b axis belongs to the 3 B3u representation, namely (b̂ + iĉ) sin(aka /2) combined with the spin part,49) and is consistent with the previously proposed non-unitary superconducting state.54, 55) This orbital one-component state is also compatible with the ultrasound measurements by Theuss et al.,19) who exclude the orbitally two-component scenario, such as {B1u , Au }, due to the accidental degeneracy of two irreducible representations in the infinitely strong SOC scheme. It is also important to examine the consistency of our results with other competing scenarios. Recent proposals of accidentally degenerate superconducting order parameters, such as the chiral B3u + iAu state15) and the B2u + iB1u or B2u + iAu states,17) allow for more complex nodal configurations. Our findings are consistent with these scenarios if multiple point 4. Summary We have performed field-angle-resolved specific-heat measurements on a clean UTe2 single crystal with T c = 2.1 K and a low residual electronic specific heat. Our results suggest the occurrence of nodal quasiparticle excitations with vF ∥ b. While the data are broadly consistent with a pointnode gap of B2u symmetry, they also allow for the possibility of line nodes confined to flat regions of the hole Fermi surfaces, where vF ∥ b holds across all kc . Our findings shed new light on the highly debated gap structure of UTe2 , offering evidence that helps elucidate its superconducting symmetry.