Spontaneous emergence of skyrmions beyond room temperature in boron-doped CrTe Authors: Jagannath Jena, Konstantin Holst, Börge Göbel, Yangkun He, Binoy Krishna Hazra, Pedram Bassirian, Katayoon Mohseni, Georg Woltersdorf, Ingrid Mertig, Holger Meyerheim, Claudia Felser, Stuart S P Parkin Published: 2026 ABSTRACT AbstractMagnetic skyrmions are vortex-like spin textures that exhibit remarkable emergent phenomena and hold great promise for applications in spintronics, magnonic devices, and neuromorphic computing. Typically, these spin textures are formed in magnetic samples through the application of magnetic fields on a ground state helical phase. However, achieving the formation of zero-field skyrmions, without the need of prior field-cooling, has remained a significant challenge. Here, using real-space Lorentz transmission electron microscopy, we directly observe the spontaneous emergence of skyrmions beyond room temperature in a centrosymmetric crystal of boron-doped CrTe without the presence of external magnetic fields. These skyrmions exhibit robust stability over a large temperature range in zero field and various specimen thicknesses. We nevertheless demonstrate that a field-cooling process can control the density and arrangement of metastable zero-field skyrmions, which is confirmed by micromagnetic simulations. The stabilization of field-free spontaneous skyrmions beyond ambient temperature holds implications for nonvolatile memory systems.Subject terms: Magnetic properties and materials, Spintronics Introduction Nanoscopic spin-textured objects, including magnetic domain walls and skyrmions, are sought after as next-generation information storage units for use in magnetic memory-storage technologies, due to their high stability, variable size, and especially their ability to be manipulated by current1–4. Skyrmions are 2D and 3D objects that typically have complex spin textures delineated by 2D or 3D boundaries (akin to domain walls) in a host magnetic material. The actual structure is realized by the competition between distinct magnetic interactions. Skyrmions have been observed in a variety of magnetic materials, primarily non-centrosymmetric systems but, in rare cases, centrosymmetric systems. In the former case, the formation of chiral skyrmions is primarily attributed to an asymmetric Dzyaloshinskii-Moriya vector exchange interaction (DMI)5,6 that appears in systems with broken inversion symmetry (examples of crystallographic point groups which fulfill this requirement are T4,7, O8, Cnv9, D2d10,11, and S412). Conversely, in centrosymmetric systems the interplay of magnetic anisotropy and dipolar interactions has been shown to allow for the generation of topologically non-trivial Bloch skyrmions with positive and negative helicities (±π/2), as well as topologically trivial type-II bubbles13–17. In many of these magnetic materials, the magnetic ground state consists of helices or stripe domains, while skyrmions are observed only in the presence of an externally applied magnetic field. Thus far, stabilizing skyrmions in zero magnetic field has been challenging and typically requires complex conditioning, such as geometrical confinement18,19, sweeping an external magnetic field from a non-zero field that allows for a skyrmion lattice state20 or cooling the sample to low temperatures in a magnetic field and then removing the field21–25. Potential applications of skyrmions are limited by the necessity of an external magnetic field, which makes such devices much more complex. It has been shown theoretically that non-centrosymmetric metallic magnets may spontaneously produce a skyrmion ground state in the absence of a magnetic field26,27. However, to date there have only been a small number of experimental examples that include monolayers of Fe and Co on Ir(111), albeit at extremely low temperatures and in which the center of symmetry is broken interfacially28,29. Recently, zero-field biskyrmions were observed in two Nd-Co-based rare-earth compounds30,31. Additionally, zero-field skyrmions have also been reported to form due to edge instabilities in the helical state under the application of current, stress, or the exchange bias effect32–34. In this study, we present the emergence of spontaneous skyrmions in the compound Cr0.9B0.1Te beyond room temperature, stabilized without any magnetic field. Real-space Lorentz transmission electron microscopy (LTEM) imaging directly visualizes these zero-field skyrmions, which coexist with short stripe domains and which exhibit remarkable stability over a wide temperature range between 100 K (the lowest accessible temperature in our microscope) and 332 K. Moreover, we demonstrate that the application of a small magnetic field leads exclusively to the formation of the skyrmion state without the presence of short stripe domains when cooling from a temperature above the critical temperature Tc of the compound down to ambient temperature. Intriguingly, the rather dense skyrmion lattice remains stable even at zero field. We elucidate the role of dipole-dipole interactions in the stabilization of these zero-field skyrmions by comparing results for different sample thicknesses. Our findings shed light on the complex behavior of skyrmion formation in this material and open new avenues for exploring their properties and potential applications. Results and discussion Single crystals of Cr0.9B0.1Te were prepared as described in reference35. X-ray diffraction (XRD) analysis of our samples reveals that the samples crystallize in the trigonal centrosymmetric space group (SGR) P3¯2/m (164) rather than the previously reported hexagonal centrosymmetric SGR P63/mmc (194). The crystal structure closely resembles the latter structure (NiAs), but we find an additional small fraction of Cr atoms (site occupancy of approximately 10%) that reside close to the tetrahedrally coordinated sites (Wyckoff site 2d(2/3, 1/3, 0.35) in SGR 164) above and below the Te atoms. Further details of the structural analysis are discussed in the Supplemental Information (Supplementary Fig. 1 and Supplementary Table 1). Out-of-plane and in-plane magnetization, M, vs. temperature, T, curves of the bulk material (Fig. 1a) were measured using a superconducting quantum interference device (SQUID) magnetometer and showed a paramagnetic to ferromagnetic transition Tc at 335 ± 1 K, which is also reflected in ac susceptibility measurements, as presented in Fig. 1b. An additional kink appears in both measurements at a lower temperature of ~138 K, that has been shown in bulk crystals of Cr0.9B0.1Te to be a spin reorientation transition temperature (TSRT), indicative of a non-collinear spin structure that forms below this temperature35. At 325 K the magnetization corresponds to 1.1 μB/f.u. reaching 2.8 μB/f.u. at 2 K (Fig. 1c). Ga+-focused ion beam (FIB) milling was used to prepare thin lamellae from the bulk crystals with the [001] axis oriented perpendicular to the lamella. Using this technique, a homogeneous specimen with a thickness of ~200 nm was fabricated. Figure 1d displays a scanning electron microscopy (SEM) image of the specimen’s surface. Figure 1e shows a selected area electron diffraction (SAED) pattern from transmission electron microscopy (TEM) studies, which confirms the [001] orientation of the specimen. Magnetic imaging from the thin specimens was performed using LTEM. LTEM imaging (i) can distinguish Bloch skyrmions7, Néel skyrmions36, topologically trivial bubbles15, antiskyrmions10 and other exotic spin textures20,37,38, (ii) has a spatial resolution of ~10 nm, and (iii) has a relatively short acquisition time compared to other imaging techniques39,40. The specimens were placed in a liquid nitrogen (LN2) based holder in which the temperature could be varied between 100 and 365 K. Crystal structure and spin textures in Cr0.9B0.1Te. a In-plane (H ∥ ab-plane) and out-of-plane (H ∥ c-axis) magnetization vs. temperature of the bulk sample for a magnetic field with strength 0.1 T. b Shows the ac susceptibility at 11 Hz and 0.1 mT ac magnetic field. c Out-of-plane magnetization vs. magnetic field of the bulk sample at various temperatures. d A thin specimen is lifted out from the bulk single crystal by using a focused ion beam (FIB) milling technique. Scanning electron microscopy SEM image shows the height and width of the specimen are 5.9 and 6.4 μm, respectively. e Selected area electron diffraction (SAED) image of the specimen. The scale bar in the diffraction pattern corresponds to reciprocal space, and the SEM image dimensions are directly labeled. f–h The upper panel shows the micromagnetically simulated spin textures of a clockwise Bloch skyrmion, counter-clockwise Bloch skyrmion, and type-II topologically trivial bubble, respectively. The lower panel shows the LTEM simulations of the corresponding textures and experimental LTEM images that are observed in a thin specimen of Cr0.9B0.1Te. The color wheel in (f) represents the in-plane magnetization distributions. The absence of a DMI in a centrosymmetric trigonal structure and the presence of dipole-dipole interactions give rise to two types of Bloch skyrmion bubbles each with a topological charge NSk=14π∬m(r)⋅∂mr∂x×∂mr∂yd2r=−1, as well as topologically trivial type-II bubbles for which NSk = 0. Micromagnetic simulations of these nano-objects, namely the clockwise (CW) Bloch skyrmion, the counter-clockwise (CCW) Bloch skyrmion, and the type-II bubble, are depicted in Fig. 1f–h, respectively. Simulations of the corresponding LTEM images and those observed experimentally match each other very well. Note that the opposite helicity of the spin textures of the CW and CCW Bloch skyrmions are reflected by the inverted LTEM contrast of the respective images. For the case of the topologically trivial type-II bubble, Bloch walls with opposite helicities are present, separated by Bloch lines (Néel-like walls). To systematically investigate the formation and stability of skyrmions, the temperature of the specimen was first increased to T = 365 K > Tc. On cooling in the absence of a magnetic field, a magnetic LTEM contrast is not observed until 325 K, where spontaneous skyrmions can be seen along with elongated objects (Fig. 2b). The LTEM images recorded on further cooling to 100 K are similar, although with increasing contrast as the magnetization systematically increases (see Fig. 2c–h). From the topological point of view, the elongated spin textures denoted by the orange and purple loops in Fig. 2f are presumably characterized by a topological charge that is the same as that of the Bloch skyrmions. For this reason, they can be considered elongated Bloch skyrmions41. A few other elongated textures have both CW and CCW Bloch walls (indicated by the orange and purple lines in a single nano-object) and are therefore topologically trivial. They are equivalent to elongated type-II bubbles. Formation of spontaneous skyrmions. a Magnetic contrast is not observable when the sample is cooled to 332 K in the absence of a magnetic field. b Zero-field Lorentz TEM magnetic contrast of skyrmions, type-II bubbles, and short spiral domains that appear at 325 K. c–h These contrasts are stable at 300, 250, 200, 130K, and 100 K, respectively. A slight change in magnetic contrast is observed at 250 K on the right corner of the LTEM image. In (f), the spin textures marked in orange or purple colors are elongated CW and CCW Bloch skyrmions, while an elongated topologically trivial bubble is marked by both orange and purple colors. The scale bar in a corresponds to 1 μm and is the same for all LTEM images. The black lines clearly visible in a are the bending contours and are present in all the LTEM images. Note that the spin textures’ contrast changes slightly in the area indicated by the red rectangles, most likely as a result of changes in magnetic anisotropy with temperature. In most previous reports, it has been demonstrated that zero-field skyrmions exist as a metastable state in a variety of materials18–22,42. This means that the skyrmions can only be stabilized at zero field after undergoing a prior field-cooling treatment. Here we show that skyrmions in boron-doped CrTe spontaneously form without the need for a magnetic field cooling protocol and that these skyrmions are stable well above room temperature. The magnetic field evolution of the spontaneous magnetic nanostructures found in Cr0.9B0.1Te was investigated by applying an out-of-plane external magnetic field. The circular and elongated objects observed at zero magnetic field are shown in Fig. 3a. The size of the elongated objects and skyrmions decreases as the magnetic field is increased, as shown in Fig. 3b, c. At 144 mT (Fig. 3d) and 162 mT (Fig. 3e), the elongated objects shrink in such a way that they form circular nano-objects that are either trivial type-II bubbles or Bloch skyrmions, depending on the initial state of the elongated object. At 192 mT (Fig. 3f), a lattice state develops, even though the periodicity is not perfect. As shown in Fig. 3g the density of nano-objects decreases, resulting in a sparse lattice state at 234 mT. With a further increase of the magnetic field to 240 mT (Fig. 3h), the field-polarized state is formed, with all moments aligned along the applied field direction so that the LTEM image contrast vanishes. Field evolution of spontaneous spin textures at 300 K and magnetic phase diagram. a Zero-field spontaneous skyrmions at 300 K. b–f At higher fields, the spiral domains transform to circular nano-objects. g, h Sparse and field-polarized magnetic textures at 234 and 240 mT, respectively. The scale bar in a corresponds to 1 μm and is the same for all LTEM images. The black lines in each LTEM image correspond to bending contours. i Magnetic phase diagram that was determined by cooling the specimen from 365 K to various temperatures in the absence of a magnetic field. The number of nano-objects per 3 × 3 μm2 area is indicated in the scale bar of (i). The red dotted arrows in the phase diagram represent the field evolution of spontaneous skyrmions at different temperatures. FP corresponds to the field polarization state. The magnetic field corresponding to the saturation magnetization at each investigated temperature in the phase diagram is indicated by green star symbols. To investigate the temperature evolution of skyrmions, the above experiment was repeated at various temperatures and a magnetic phase diagram was constructed, as depicted in Fig. 3i. The color code in the diagram represents the density of the observed nano-objects, including Bloch skyrmions of both helicities and topologically trivial bubbles. As evidenced in the phase diagram, the field stability region of the skyrmions increases with decreasing temperature. The field evolution of the spin textures at 100 K is shown in Supplementary Fig. 2, which shows that the field stability region is approximately twice as large as at 300 K. So far, all experiments have been conducted following a zero-field cooling protocol for the initial formation of the magnetic state. Next, we perform field-cooling (FC) experiments to allow for different magnetic configurations that may arise due to the metastability of the nano-objects in our sample (Fig. 4). Initially, the specimen was heated to T = 365 K > Tc, followed by the application of a finite magnetic field at that temperature. The specimen is then cooled to ambient temperature while maintaining the constant magnetic field. We repeated this FC method using different strengths of the magnetic field. Figure 4a–f show the LTEM contrasts after FC to ambient temperature under various magnetic field strengths. At 6 mT (Fig. 4a) the result is similar to our previous observations at 0 mT, with both circular and elongated nano-objects. When a higher magnetic field strength is applied during the FC process, the density of nano-objects becomes larger. The elongated objects disappear above 15 mT (Fig. 4b, c). In the latter case, the skyrmions and topologically trivial bubbles are smaller. With further increasing field during the FC process, the density of all nano-objects decreases, as observed for 66 mT (Fig. 4d) and 81 mT (Fig. 4e), and eventually, a field-polarized state is observed at 120 mT (Fig. 4f). Besides these findings, we noticed that a greater proportion of topologically trivial bubbles are formed at higher field strengths whereas the proportion of Bloch skyrmions is higher at lower field strengths during FC. Stability after different field-cooling protocols. The specimen is cooled down from 365 to 300 K in the presence of various magnetic fields. a-f Upper panels: LTEM contrasts of the specimen after magnetic fields of 6, 15, 51, 66, 81, or 120 mT have been applied for field cooling, respectively. g–l Corresponding LTEM contrasts acquired after reducing the magnetic field to zero. m–r Corresponding LTEM contrasts taken after increasing the magnetic field to 192 mT. The scale bar in the upper panel of (a) corresponds to 1 μm and is the same for all LTEM images. Bending contours of black lines are clearly visible in each LTEM image. s Number of nano-objects found in the area of 3 × 3 μm for zero-field cooling (black curve) and field cooling processes (values as indicated). The open stars correspond to fields for which field cooling was performed, and then the field was reduced. The filled stars correspond to the case where the field was increased starting from zero field. After the individual magnetic configurations have been stabilized under the above-explained FC protocol with distinct magnetic field strengths, we compare how they change when the magnetic field is reduced to zero after their formation (Fig. 4g–l). For the configurations that have initially been formed under FC between 6 and 66 mT, the nano-objects remain stable (Fig. 4g–j), and we observe no notable change in the magnetic textures. In these cases, the skyrmion density is comparably high. However, for the configuration that was field-cooled at 81 mT, the number of non-collinear objects in the sample is lower and a reduction of the magnetic field to zero leads to size changes of several magnetic textures (Fig. 4k). For the ferromagnetic state that appears under field cooling at 120 mT, the region is filled only with spirals in the absence of a magnetic field (Fig. 4l). These results point out the importance of the experimental protocol for the formation of non-collinear spin textures. All the presented images in Fig. 4g–l have been recorded under exactly the same conditions, but the sample hosts different numbers and types of spin textures caused by the distinct experimental histories. While applying a magnetic field during the cooling process from 365 K to room temperature is generally not required for the formation of bubbles and skyrmions, their nucleation can be fostered, and their density increased when a finite magnetic field is applied. The ideal FC field for that was ~51 mT in our experiments. Continuing from the distinct zero-field states shown in Fig. 4g–l, the magnetic field is now increased to 192 mT (Fig. 4m–r). In all cases, the number of stabilized nano-objects remains almost the same, and the size of the nano-objects decreases compared to the state at 0 mT. Still, major changes are observed, most importantly for the measurement that started with FC at 81 mT (Fig. 4e). Here, the shrinking of the textures causes the initially elongated objects to form almost perfectly circular skyrmionic bubbles. As before, the various configurations recorded at the same field of 192 mT are very different from each other due to the experimental history. Figure 4s summarizes and generalizes the above-discussed results. The plot depicts the number of nano-objects (skyrmions and bubbles) that we have observed for different values of the applied magnetic field. The individual curves differ only by their starting configuration, which was generated under different magnetic field strengths applied during field cooling. These configurations correspond to Fig. 4a–f, as indicated. As mentioned before, the density of nano-objects shows an increasing trend with the magnetic field that was applied during field-cooling up to a certain strength and drops for higher values. For the state that was stabilized under a field of 51 mT, at which we see the largest number of nano-objects, many of those have 6 neighbors. A nearly periodic, hexagonal lattice is formed (see Supplementary Fig. 3). However, the further we go away from this field during FC, the fewer nano-objects we observe and the more random the ordering appears. Once the saturation state is achieved, reducing the magnetic field typically leads to the formation of a labyrinthine domain state, occasionally accompanied by a few isolated nano-objects embedded within it. This behavior is consistent across all low temperatures, except near Tc. Near Tc, we observe a distinct response: nano-objects emerge directly during the field reduction from the polarized (saturated) state, as illustrated in Supplementary Fig. 4. The experimental results indicate a manifold of allowed metastable configurations. The stabilized magnetic state can be controlled by the precise experimental protocol. In the following, we model this process by micromagnetic simulations (at zero temperature) and show that the stabilized configuration strongly depends on the starting configuration, which acts as a seed. Note that periodic boundary conditions are imposed. The seed emulates the different field-cooling conditions from Fig. 4. We have used the specific material parameters of the measured sample. Technical details about the simulations are presented in the Methods section. As shown in Fig. 2, zero-field cooling leads to a mixed magnetic state consisting of circular and elongated skyrmionic bubbles of both helicities as well as topologically trivial bubbles. This process can be emulated in micromagnetic simulations by starting the simulation with a seed that contains only a few skyrmions. For example (Fig. 5a), nine skyrmions turn into a metastable state exhibiting five skyrmionic bubbles and four elongated magnetic textures. When the skyrmion number in the seed is decreased to four (Fig. 5b), the result is a configuration consisting of four elongated textures. This is similar to the experimental scenario of field-cooling under a rather large magnetic field (like in Fig. 4e, FC 81 mT). Once the field has been turned off, mostly helices are present. Various metastable spin textures in micromagnetic simulations. a Starting from 9 Bloch skyrmions leads to a metastable configuration of 5 deformed skyrmions and 4 elongated skyrmions that form short stripe domains. b Four circular Bloch skyrmions elongate and form short stripe domains. c Sixteen Bloch skyrmions with helicities ±π/2 form a periodic skyrmion lattice. d Starting from 36 Bloch skyrmions leads to a merging until only 20 Bloch skyrmions of various sizes form a metastable configuration. Periodic-boundary conditions have been used. For more details, see “Methods”. The scale bar in (b) corresponds to 1 µm and is the same for all images. In the experiments, a dense skyrmion lattice can be stabilized under field-cooling at 51 mT as presented in Fig. 4c. In the simulation, this state is also stable; see Fig. 5c. If the seed is chosen to be a periodic lattice of 16 skyrmions, the configuration remains metastable, and we even observe both skyrmion helicities. Topologically trivial bubbles are unstable, which could be related to the fact that our simulations have been carried out without taking thermal effects into account. Note that in the experiments, skyrmions become more stable than topologically trivial bubbles at low temperatures as well. When we start with an even higher number of skyrmions in the simulations, here 36, the individual skyrmions are too small to be stable. Some of these skyrmions spontaneously disappear but others merge, so that the number of skyrmions in the final stable configuration is reduced (here to 20) (Fig. 5d). This state is not periodic and the individual skyrmions have different sizes and shapes, just like the experimental results e.g., in Fig. 4. The magnetic texture is trapped in a local energy minimum and cannot achieve the periodic lattice state. Lastly, we investigate the thickness dependence of the skyrmions and topologically trivial bubbles. We carried out similar experiments on a second lamella that has two regions with thicknesses of 160 and 270 nm (Fig. 6a) compared to the thickness of 200 nm that was used before. In this lamella, we also observe the emergence of spontaneous skyrmions at various temperatures at and beyond room temperature (Fig. 6b–e). Therefore, it is reasonable to assert that the formation of spontaneous skyrmion spin textures is inherent in the material. Interestingly, the thicker region exhibits spontaneous magnetic contrast even at 332 K, whereas the thinner region does not. This could be caused by a slightly different energy landscape due to an increased importance of the dipolar interactions. For this reason, we also see that the size of the skyrmions increases as the thickness of the specimen increases (See Table 2 in Supplementary Information.). Furthermore, we find that the field stability region of nano-objects for thicker specimens is larger than for thinner ones. This is presented in the Supplementary Fig. 5. However, another possibility for why we do not see any magnetic contrast in the thinner region at 332 K could be the smaller magnetic induction compared to the thicker region, which deflects the transmitted electron beam less after passing through the specimen. Therefore, even though the compound has a very low moment close to Tc, skyrmionic spin textures can still be observed in the thicker region. In agreement with this argument, the thinner region generally shows a weaker magnetic contrast than the thicker region (Fig. 6c–e). Stability of zero-field spin textures for thinner and thicker specimen. a SEM images showing the side view (surface in the left image) and top view (different thicknesses in the right image) of the specimen. b The LTEM overview of the specimen at 300 K and 0 mT. Magnetic contrast is not visible in the upper corner of the specimen due to bending contours. The variation in image contrast originates from thickness variations within the TEM lamella. Reduced contrast in the thinner region results from a smaller integrated magnetic induction, whereas reduced contrast in the thicker region arises from decreased electron transmission. c–e LTEM contrasts captured at 332, 325, and 300 K, respectively, for thicker (270 nm, left side) and thinner (160 nm, right side) regions of the specimen. These images are taken from the same regions marked in yellow and orange. The scale bar for the left side LTEM image in (c) corresponds to 500 nm and is the same for the images in (c–e). f LTEM contrast for the thicker region at 300 K after field cooling the specimen from 365 K at a constant magnetic field of 12 mT. g, h The specimen was consequently cooled to 200 and 100 K, keeping the magnetic field constant. Differently colored arrows in (f, g) point to the same textures at different temperatures that have transformed (red: from trivial bubble to CW Bloch skyrmion, green: from trivial bubble to CCW Bloch skyrmion, blue: from CW Bloch skyrmion to CCW, purple: from CCW Bloch skyrmion to CW). The scale bar in (h) corresponds to 500 nm and is the same for the LTEM images in (f, g). In the thicker region of the sample, we also observe conversion processes when the specimen was cooled to a low temperature from 300 K under a constant magnetic field. As seen in Fig. 6f–h, a few trivial bubbles convert to Bloch skyrmions and some Bloch skyrmions change their helicity (CW to CCW and vice versa). The red and green arrows mark the conversion of trivial bubbles to CW Bloch skyrmions and CCW Bloch skyrmions, respectively, when cooling from 300 to 200 K. Additionally, the blue arrows highlight the conversion of CW Bloch skyrmions at 300 K to CCW Bloch skyrmions at 200 K, whereas the purple arrows indicate the reverse conversion process. Moreover, a few of the nano-objects vanish at 200 K, causing the neighboring objects to expand in size (Supplementary Fig. 6). As discussed above, the magnetic objects try to fill the available space as much as possible. There were no such changes in spin textures when the specimen was cooled from 200 to 100 K. Conclusion In summary, we have observed the spontaneous emergence of zero-field Bloch-like skyrmions and type-II bubbles over a wide temperature range from 100 K to beyond room temperature at 332 K, in the absence of any magnetic field. We have shown that these objects are stabilized by dipole-dipole interactions in Cr0.9B0.1Te, a centrosymmetric material and, thus, without any intrinsic DMI. The skyrmions coexist with a few topologically trivial bubbles and elongated objects, where the detailed magnetic configuration depends strongly on the experimental protocol used. The density of the skyrmions and whether they form a lattice state depends on the field-cooling procedure. Unlike DMI-stabilized skyrmions, these skyrmions and bubbles are rather flexible, and their density can be varied greatly. The field-cooling procedure can be used to control the nucleation rate of the skyrmions. Depending on the number of nano-objects in the sample, they are circular in the preferred magnetic field range, but they elongate and fill up all the available space once the field is absent because they try to turn into spiral domains. This is confirmed by micromagnetic simulations by means of different seed configurations that lead to different metastable magnetic configurations once relaxed. The discovery of spontaneous skyrmions in the crystal Cr0.9B0.1Te occurring without any external field or geometric constraint contributes significantly to our understanding of skyrmion physics. This new understanding will assist the analysis of the 3D magnetization of skyrmions without the need for complex experimental setups because these imaging techniques will not be disturbed by the need for magnetic fields43. Moreover, the spontaneous coexistence of skyrmions with elongated nano-objects might serve as a reservoir with enhanced nonlinear functionalities in the realm of reservoir computing. All desirable features for the application of skyrmions, such as a wide temperature range, high density, zero magnetic field, and thickness-independent stability, were simultaneously observed in the same material in our study, widening the applicability and opening exciting avenues for skyrmion physics. Methods Lamella preparation Single-crystalline Cr0.9B0.1Te specimens were produced from the [001] oriented surface of the bulk crystal by using a Ga+ ion-based dual-beam focused ion beam (FIB) system [FEI Nova Nanolab 600 SEM/FIB]. The FIB is operated at an accelerating voltage of 2–30 kV. The final smooth polishing of the specimen is carried out at 2 kV and 4.4 pA. Scanning electron microscope (SEM) imaging inside the FIB system was used to measure the thickness of the specimens under a zero-tilt stage condition, and the dimensions of the prepared specimen were determined from SEM imaging at 52° normal to the surface. X-ray diffraction The structural analysis of the Cr0.9B0.1Te crystal was carried out using a six-circle diffractometer and a Gallium-Jet X-ray source (λ = 1.3414 Å) operated at 70 keV and 200 W power. The beam was focused onto the sample by Montel optics providing a highly collimated beam of 100 μm height and 2 mm width. Integrated reflection intensities were collected using transverse phi-scans of the sample and a two-dimensional pixel detector. Transmission electron microscopy and analysis of the magnetic nanostructure The SAED and LTEM measurements were performed in a high-resolution transmission electron microscope [JEOL JEM-F200]. A 100 μm selected area aperture was used to capture the diffraction pattern of the single crystalline specimens. A beam stopper was used to block the high-intensity (000) central diffraction spot. Magnetic domain contrasts were acquired in Fresnel mode (LTEM mode) with a negligible remnant magnetic field (<2 mT). The magnetic field was varied along the microscope axis by exciting the objective lens. A double-tilt liquid nitrogen holder was used to adjust the specimen’s temperature, and a temperature controller was used to maintain the temperature of the specimens at varied levels. The specimens were not tilted at any point during cooling at any temperature, magnetic imaging, or applied magnetic field. The over-focus value for the LTEM imaging is ~0.7–0.9 mm. Micromagnetic simulations We have used the GPU-accelerated micromagnetic code mumax3 to stabilize the different magnetic configurations in Fig. 5. The corresponding seed magnetizations (Supplementary Fig. 7) have been constructed from circular Bloch skyrmions that are positioned in a sample with periodic boundary conditions along x and y to emulate an extended sample. The simulated volume has the dimensions 2000nm×1800nm×200nm in Fig. 5b, c. To avoid finite size effects, the dimensions in Fig. 5a, d are 2000nm×2000nm×200nm. A cell size of 5×5×5nm3 was used. Values of saturation magnetization of Ms=3.42×105A/m, exchange A=2.2×10−11J/m, and anisotropy Kz=0.4×105J/m3 were used. Compared to Liu et al.44, we have measured a significantly smaller uniaxial anisotropy Kz (their value Kz=0.79×105J/m3). This leads to a larger bubble size and especially to larger domain walls in the bubbles in both the experiment and the micromagnetic simulations. For the relaxation, we have used the “relax” command based on energy minimization and an RK23 method (Runge–Kutta method), followed by a 50 ns propagation according to the Landau–Lifshitz Gilbert equation and an RK45 method. The seed configurations are shown in the supplementary material (Supplementary Fig. 7). Supplementary information SI text and figures Transparent Peer Review file Article File-pdf LaTeX Supplementary File Acknowledgements We acknowledge funding from the European Research Council (ERC) under the European Union’s Horizon 2020 research and innovation program (grant agreement SORBET No. 670166) and the Deutsche Forschungsgemeinschaft (DFG, German Research Foundation)—Project number 403505322 under SPP2137. C.F. acknowledges funding by the DFG under SPP 2137 (Project number 403502666). I.M. and B.G. acknowledge funding by the EIC Pathfinder OPEN Grant No. 101129641, “Orbital Engineering for Innovative Electronics”. Author contributions J.J. and S.S.P.P. conceived the study. J.J. carried out the Lorentz TEM observations, and J.J. and K.H. analyzed Lorentz TEM data. The bulk material was synthesized by Y.H. and C.F. J.J. and K.H. prepared the lamellae of different thicknesses. B.G. and K.H. performed the micromagnetic simulations. P.B., I.M., and G.W. helped with the analysis of the manuscript results. J.J. and B.K.H. performed the magnetization measurements. H.M. and K.M. contributed to the crystal structure analysis. J.J., K.H., B.G., and S.S.P.P. wrote the manuscript with substantial contributions from all authors. S.S.P.P. supervised the project. Peer review Peer review information Communications Materials thanks the anonymous reviewers for their contribution to the peer review of this work. A peer review file is available. Funding Open Access funding enabled and organized by Projekt DEAL. Data availability The data that support the findings of this study are available from the corresponding authors upon reasonable request. Competing interests The authors declare no competing interests. Footnotes Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations. These authors contributed equally: Jagannath Jena, Konstantin Holst, Börge Göbel. Contributor Information Jagannath Jena, Email: jjena@anl.gov. Stuart S. P. Parkin, Email: stuart.parkin@mpi-halle.mpg.de Supplementary information The online version contains supplementary material available at 10.1038/s43246-026-01336-1. References Associated Data Supplementary Materials SI text and figures Transparent Peer Review file Article File-pdf LaTeX Supplementary File Data Availability Statement The data that support the findings of this study are available from the corresponding authors upon reasonable request.