WorksPhysics and Astronomy
Magnetic phase diagram of cuprates and universal scaling laws
Abstract
In this article we consider the magnetic field phase diagram of hole-doped high- cuprates, which has been given much less attention than the temperature diagram. In the framework of the pairon model, we show that the two characteristic energies, the pair binding energy (the gap ) and the condensation energy () resulting from pair correlations, give rise to two major magnetic fields, the upper critical field and a second field, , associated with the pseudogap (PG). The latter implies a second length scale in addition to the coherence length, characteristic of incoherent pairs. Universal scaling laws for both and are derived: scales with the critical temperature, T/K, in agreement with many experiments, and has a similar scaling with respect to . Finally, Fermi arcs centered on the nodal directions are predicted to appear as a function of magnetic field, an effect testable experimentally.
Résumé
*Corresponding author : [email protected]
PACS numbers: 74.72.h,74.20.Mn,74.20.Fg
Introduction
Establishing the magnetic phase diagram as a function of doping is a key step to grasp the fundamental mechanisms underlying superconductivity in cuprates. Under an applied magnetic field, type II superconductors satisfy general properties. Vortices appear above the lower critical field, , and the typical size of the vortex core is the superconducting (SC) coherence length . The latter is directly related to the upper critical field, , through the relation
where is the magnetic flux quantum.
In cuprates the phase coherence length is in the nanometer range so that the upper critical field is often too large to be measured, in which case its estimation can be made using indirect extrapolation methods. Such is the case, for instance, for YBaCuO (YBCO) or BiSrCaCuO (BSCCO). This experimental limitation might explain some of the discrepancies for obtained in different experiments.
The upper critical field has been deduced from the specific heat by different groups [1, 2] on LaSrCuO (LSCO) for different hole concentrations () from underdoped to overdoped sides of the critical dome. The authors measured the dependence of the coefficient as a function of magnetic field, from which can be obtained. Their results indicate that follows a dome as a function of , and hence follows the critical temperature. These findings are in qualitative agreement with the high-field measurements of the resistivity [3]. On the other hand, there is an apparent contradiction with experiments by Kato et al. Kato et al. [4], who estimate the upper critical field of BiSrCaCuO from the measured irreversibility field. In this case, qualitatively follows a dome for small carrier concentrations, but then increases monotonically to high values in the overdoped regime. Therefore, in order to address these issues, it is important to clarify the key parameters, i.e. the characteristic energies, lengths, and magnetic fields of high- cuprates.
More than 30 years after the discovery of cuprates by Bednorz and Müller [5], key questions such as the pairing ‘glue’, the nature of the pseudogap, and the role of the gap parameter, remain open issues. Above , contrary to conventional superconductors, the normal metallic state is not recovered. Instead, the pseudogap (PG) state is found, characterized by a lowering of the quasiparticle DOS, which persists up to the higher temperature [6].
In the BCS theory, the energy gap is proportional to the critical temperature: . The situation is different in cuprates since the amplitude of the gap is not directly related to . A gap of the same order of magnitude persists above [7] as well as within the vortex core, see Figure 1, panel (a). This is confirmed by scanning tunneling spectroscopy (STS) measurements in disordered BSCCO thin films [8]. In the SC state, the quasiparticle spectrum displays the peak–dip structure. Crossing to the non-SC region, the coherence peaks disappear while the gap is preserved having roughly the same magnitude, see Figure 1, panel (d). A similar transition is also observed as a function of temperature [7]. Therefore, since a gap exists even when the superfluid density vanishes, it cannot be the order parameter.

Figure 1. Strong experimental evidence of the pseudogap (PG) state obtained when SC coherence is lost at low temperatures. In contrast to the conventional case, this PG state consists of incoherent pairons, even at the local scale[18]. a) STS spectrum taken at low temperature in the SC state as well as within the vortex core (adapted from [12]). b) STS spectra taken along a line crossing a vortex (adapted from [13]). c) Theoretical spectra (solid blue line) in the SC state and in the incoherent pseudogap state, red dashed line (adapted from [14, 15, 16]). d) STS spectra along a line crossing from an ordered region to a disordered region (adapted from [8]). All panels correspond to BSCCO(2212).
This conclusion is confirmed by the dependence of the energy gap with doping. In fact, the gap, which is directly measured by tunneling or photoemission spectroscopy, decreases linearly as a function of carrier density [9, 10, 11]. In the underdoped regime, has the opposite behavior with respect to the energy gap since it increases with doping. Two undecided questions thus emerge from these considerations :
What is the order parameter in cuprates ?
What is the magnetic phase diagram and its connection to the temperature phase diagram ?
In this article, we establish important physical parameters of hole-doped cuprates in the framework of the pairon model. Based on a two–fluid Ginzburg-Landau approach, we obtain the expressions for the characteristic lengths and magnetic fields. This allows to deduce the magnetic phase diagram of cuprates, and to compare it with the analogous temperature diagram.
Characteristic energy scales of cuprates
In order to answer the above questions, we consider the pairon model wherein the SC state is the result of the condensation of preformed pairs due to their mutual interaction [17, 18, 19]. Once SC coherence is lost due to a magnetic field or rising temperature, the system becomes a ‘glassy’ PG state of incoherent pairons, such as in the vortex core. The corresponding phase diagram for hole-doped cuprates can then be described by two energy scales : the pairing energy gap and the condensation (or correlation energy) .
The meanings of and in the pairon model are distinct. Pairons are bound holes on adjacent copper sites which form below as a result of their local antiferromagnetic environment on the scale of , the AF correlation length [18]. They condense into a superconducting state at the critical temperature (), forming a collective pairon state. The energy (per pairon) in the pseudogap state and in the SC state are respectively :
It is important to note that the condensation energy is due to correlations between pairons. In a sense it is a ‘hidden’ term since, contrary to conventional superconductors, it does not correspond to a binding energy and is not measured as a usual gap in photoemission or scanning tunneling spectroscopy. However the coherence energy, , is present in the quasiparticle spectrum for energies just above the coherence peak characterized by a sharp dip, see Figure 1(c). The precise analysis of the quasiparticle spectra using a gap function reveals quantitatively both energies and [15, 16, 17].
We have previously shown that both and are proportional to the same fundamental physical parameter, the effective exchange energy in the CuO plane. Note that is material dependent since it varies as a function of the c-axis coupling between CuO planes. This can explain the variations of between different materials.
As a result of topological constraints of holes in the CuO plane [19], the following relations are derived for and as a function of the hole concentration :
Here is the reduced density, with and . Thus, conveniently ranges from 0 to 1, from the start to end of the dome. Moreover, the energy and temperature scales are linked by the relations [18] :
Considering the available data on BSCCO(2212), the numerical factor does have some uncertainty : . The simplest ratio of the two energies therefore leads to :
For LSCO this ratio is slightly lower () due to the relatively smaller . As shown in a previous article [20], this ratio also determines the size of the Fermi arcs at the critical temperature.
Characteristic lengths
A two-fluid model is a convenient way to derive the appropriate length scales. In order to describe the two different fluids, the superfluid (the coherent pairons of the condensate) and the fluid of incoherent pairons (pairons excited out of the condensate), we consider two coupled Ginzburg-Landau equations in zero applied field:
Here is the condensate density and is the excited pairs density. For example, qualitatively, the condensate density gives rise to the typical STS spectrum outside the vortex, whereas aptly describes the local density within the core (see Figure 1 (b)).
Neglecting pair-breaking below and phase fluctuations above , the conditions that for and for should be fulfilled. The variation of the free energy between the normal state and superconducting state (assuming ) is
In addition, we have . This allows to compare to the expression of expected in the pairon model, from which we deduce that . Including pair-breaking below , relevant to the highly-overdoped regime, gives the same result.
We therefore obtain the two length scales , with , associated with the two energies and :
Here is the length associated with the pairing gap while is the coherence length associated with superconducting phase coherence.
A third important length scale is the antiferromagnetic correlation length and its dependence on the hole carriers, . In a simple approach [18], we proposed that is determined by the mean pairon-pairon distance in the CuO plane:
where is the in-plane Cu-Cu distance. This formula is in qualitative agreement with neutron experiments [21]. In this picture, a given pairon is surrounded by an area, a Voronoi cell, corresponding to a well-defined AF magnetic order on the scale of . In [18], we showed that the pairon binding energy, , is proportional to the Voronoi cell area, providing a simple explanation for its -dependence.
It is instructive to see that, after a little algebra, two of the length scales and are connected by a simple equation:
where is the value of extrapolated to , and . The above equation can be represented by a hyperbola (solid black line Figure 2, right panel). The third length, , can be conveniently expressed as:

Figure 2. (Color online) Plot of the coherence length , the pair correlation length , and the antiferromagnetic correlation length , according to the model. a) All three lengths plotted as a function of carrier concentration. b) , plotted as a function of . We note the special points c’ and c, at the values , and , respectively. The plots are appropriate for BSCCO(2212).
The variation of the three length scales , and as a function of carrier concentration are shown in Figure 2, left panel. As seen in the figure, there is a striking difference in the behavior of and . In the highly underdoped regime , evidently the pairons behave dominantly as bosons. However, in the overdoped regime, i.e. towards the end of the -dome, as , so the two length scales approach each other asymptotically. At the same time, as can be seen using eqns. (3) and (4), close to the critical point or , the associated energy scales and , also approach each other, and vanish identically in the limit . The overdoped side of the dome thus recalls the BCS case where the coherence length is determined by the energy gap. However, it would be misleading to conclude that one recovers the BCS mechanism in this regime. Indeed, the underdoped and overdoped regimes clearly correspond to different limits, given by the ratio . However, the mechanism at the heart of the cuprate SC coherence is imposed by the conjunction of the geometrical constraints of the CuO plane and the finite size of hole-pairs [19]. Detailed considerations of the model confirm that pairons, i.e. two holes bound on neighboring copper sites in an antiferromagnetic environment, are the fundamental quantum objects across the -dome.
To conclude this discussion, we note that two special points in the phase diagram exist : c’ and c (respectively, at the concentrations and ), see Figure 2. These two points correspond to the common tangent of with and the intersection of with , respectively. We see the further condition that for all , and that c’ (at ) is an evident transition point involving the characteristic length . The latter phenomenon will be tackled in a future report.
Characteristic magnetic fields of cuprates
The coherence length , is the typical size of the vortex core and it is associated with the upper critical field . Similarly, is associated with a higher field reflecting the pseudogap state. While is the field necessary to destroy SC phase coherence, we shall show that pairons theoretically survive up to .
From the two length scales, one can deduce the expressions of the two important magnetic fields using the relations :
Replacing the expressions for and , we obtain :
One first important consequence from Eq. (18) is that the upper critical field must follow the superconducting dome. This is in good agreement with the findings of Wang and Wen [1], as shown in Figure 3.

Figure 3. (Color online) Upper critical field, , for LSCO as a function of carrier density. Experimental values (red points) are taken from Wang et H.-H. Wen [1]. Dashed blue line : theoretical prediction from the model.
One can therefore deduce important scaling relations. Since the upper critical field in the pairon model is proportional to the coherence energy , we have :
where T/meV. Using , it follows that :
This relation should hold for any hole-doped cuprate and for arbitrary carrier concentrations within the SC range. In Figure 4, we have plotted the upper critical field for several materials and doping values. The agreement confirms the linear relation between and . We note that a wide variety of conventional superconductors follow a similar trend (inset of Figure 4), however in this case each material has a fixed carrier concentration, in contrast to cuprates. Finally, using Eqs. (5) and (19), the model predicts a similar scaling of the pairing field with : T/K.

Figure 4. (Color online) Upper critical field, , as a function of the critical temperature, , for different cuprate materials and doping values. Inset : Similar plot for conventional superconductors A to H, follows a similar line with T/K. In the main plot : KE : Krusin Elbaum et al. BSCCO [22], K : Kugler et al. BSCCO [13], M : Maggio Aprile et al. YBCO [23], W : Wang et Wen LSCO [1], G : Grisonnanche et al. YBCO [3], G2 : Grisonnanche et al. Tl2201 [3]. Details for the inset : A : BaBi [24], B : NbSe [25, 26], C : MgCNi [27], D : NbTi [28], E : VGa [29], F : VSi [30, 31], G : NbSn [30], H : NbAl [32], I : RbC [33].
Magnetic phase diagram and discussion
The magnetic phase diagram thus appears to be very analogous to the temperature phase diagram. Pairons form below the pseudogap field and become coherent below , see Figure 5. This is similar to the phase diagram proposed by Krusin-Elbaum et al. [22], based on high magnetic field c-axis resistivity measurements. The striking analogy between the temperature and magnetic phase diagrams suggests that the magnetic field and the temperature produce similar effects. Phase coherence is destroyed either by raising the temperature above , or by applying a magnetic field stronger than , as in conventional superconductors. However, pairons survive in cuprates above up to and, as our work confirms, above up to . The interpretation of as the pair-breaking field is thus reasonable, whereas breaks the phase coherence, in clear agreement with experiment (Figure 1).

Figure 5. (Color online) Magnetic phase diagram showing the two main characteristic fields, the upper critical field and the field associated with pseudogap (solid black line). Dashed green line: deduced using the estimated value of the coherence length from the vortex core in BSCCO (Figure 1, panel (b)). Solid green line: of BSCCO using the theoretical value for . Solid blue line: predicted for LSCO. Dashed black line: upper limit for in the model (associated with the length scale ). Note that pairons form below the pseudogap magnetic field, , and become coherent below .
A unique property of cuprates is the existence of Fermi arcs of ‘normal’ electrons centered around the nodal directions at finite (see [6] and Refs. therein). In our model hole pairons are associated with a continuum of electron Cooper pairs in momentum space. Four Fermi arcs then appear due to the weaker binding energy of Cooper pairs with momenta close to the nodal directions [20]. As shown using ARPES experiments [6], a given Fermi arc is very small at low temperature, in agreement with d-wave pairing. Then, as , the Fermi arcs grow monotonically with temperature, at a rate that is -dependent. The complete Fermi surface is recovered at the higher temperature [20].
To quantify the Fermi arc size using simple arguments, we recall that the d-wave pair potential is well approximated by: where is the momentum angle measured from the antinodal direction. Thus, at some finite temperature , pairs whose energy satisfies will decay into normal electron states around the node. The Fermi arcs do indeed increase with temperature, an effect which is more pronounced for higher carrier concentrations. In particular, at we obtain the critical angle which characterizes the width of the Fermi arc:
This critical angle has been directly measured by Hashimoto et al. [6] where, for near optimal doping, the above ratio is about giving . (As previously mentioned, the simple ratio above, valid for BSCCO-2212, would be slightly smaller for LSCO by about 20%.)
The same physical effect should also occur in a magnetic field. Indeed, using eqns. (18) and (19) we obtain the analogous relation for the critical angle:
Again, the complete Fermi surface is only recovered at the higher pairing field, . The existence of such Fermi arcs for the lower field, , can be considered a critical test for the pairon model.
Conclusion
In conclusion, we have derived expressions for the length scales and magnetic fields of hole-doped cuprates. A two-fluid model, using two coupled Ginzburg-Landau equations, describes the condensate (the coherent pairs) and the incoherent pairs (pairs excited out of the condensate). The two energy scales describing cuprates, the energy gap, , and the condensation energy, , give rise to two characteristic magnetic fields, the upper critical field, , and the pseudogap field, . The magnetic phase diagram is therefore very analogous to the temperature phase diagram. It appears that the same critical point, , at the end of the dome, also exists in the magnetic diagram.
In this work, we have established a universal scaling law for as a function of the critical temperature, T/K, in quantitative agreement with experiments, which should be valid for all cuprates on the hole-doped side, and for any carrier concentration. A similar scaling law is predicted for the pseudogap field, , but with respect to .
Instead of a single coherence length , the model invokes a second length, , associated with the pairing gap. Their dependence on the carrier concentration, from underdoped to overdoped sides of the phase diagram, offers unique insights into the cuprate mechanism. The present theory can be used to express the thermodynamic field, the condensation energy, the lower critical field , and the London penetration depth , which we leave for a future report.
Finally, we propose that Fermi arcs should also exist as a function of magnetic field, a prediction that can be used as a strong experimental test of the model.
Acknowledgements
The authors gratefully acknowledge discussions with Dr. Hiroshi Eisaki, Dr. Shigeyuki Ishida (AIST, Tsukuba), and Prof. Atsushi Fujimori (Tokyo University).
A.M. and W.S. acknowledge partial support from the French National Research Agency (ANR), project ‘Superstrong’ under contract no. ANR-22-CE30-0010 (Principal Investigator : Andrea Gauzzi).
W.S. is grateful for continual support of the RIIS Institute of Okayama University, Japan, to Prof. Takayoshi Yokoya (host), and his ‘visiting professor’ status while on leave from SU.
Références
References
- [1]Y. Wang and H.-H. Wen, Doping dependence of the upper critical field in La₂₋ₓSrₓCuO₄ from specific heat, Europhys. Lett., 81, 57007 (2008).
- [2]H. H. Wen, H. P. Yang, S. L. Li, X. H. Zeng, A. A. Soukassian, W. D. Si and X. X. Xi, Hole doping dependence of the coherence length in La₂₋ₓSrₓCuO₄ thin films, Europhys. Lett., 64, 790 (2003).
- [3]G. Grissonnanche, O. Cyr-Choinière, F. Laliberté, S. René de Cotret, A. Juneau-Fecteau, S. Dufour-Beauséjour, M. -È. Delage, D. LeBoeuf, J. Chang, B. J. Ramshaw, D. A. Bonn, W. N. Hardy, R. Liang, S. Adachi, N. E. Hussey, B. Vignolle, C. Proust, M. Sutherland, S. Krämer, J. -H. Park, D. Graf, N. Doiron-Leyraud and Louis Taillefer, Direct measurement of the upper critical field in cuprate superconductors, Nature Communications 5, Article number : 3280 (2014).DOI
- [4]Junichiro Kato, Shigeyuki Ishida, Tatsunori Okada, Shungo Nakagawa, Yutaro Mino, Yoichi Higashi, Takanari Kashiwagi , Satoshi Awaji , Akira Iyo , Hiraku Ogino, Yasunori Mawatari, Nao Takeshita, Yoshiyuki Yoshida, Hiroshi Eisaki, and Taichiro Nishio, Doping dependence of upper critical field of high-Tc cuprate Bi₂₊ₓSr₂₋ₓCaCu₂O₈₊δ estimated from irreversibility field at zero temperature, J. Phys. Soc. Jpn. 93, 104705 (2024).DOI
- [5]J. G. Bednorz, K. A. Müller, Possible high Tc superconductivity in the Ba-La-Cu-O system , Zeitschrift für Physik B Condensed Matter 64, 189 (1986).
- [6]Makoto Hashimoto, Inna M. Vishik, Rui-Hua He, Thomas P. Devereaux and Zhi-Xun Shen, Energy gaps in high-transition-temperature cuprate superconductors, Nature Physics 10, 483 (2014).
- [7]Ch. Renner, B. Revaz, J.-Y. Genoud, K. Kadowaki, and Ø. Fischer, Pseudogap precursor of the superconducting gap in under- and overdoped Bi₂Sr₂CaCu₂O₈₊δ, Phys. Rev. Lett. 80, 149 (1998).
- [8]T. Cren, D. Roditchev, W. Sacks, J. Klein, J.-B. Moussy, C. Deville-Cavellin, and M. Laguës, Influence of disorder on the local density of states in high-Tc superconducting thin films, Physical Review Letters 84, 147 (2000).
- [9]Tohru Nakano, Naoki Momono, Migaku Oda, and Masayuki Ido, Correlation between the doping dependences of superconducting gap magnitude 2Δ₀ and pseudogap temperature T* in high-Tc cuprates , J. Phys. Soc. Jpn. 67, 2622-2625 (1998).
- [10]S. Hüfner, M. A. Hossain, A Damascelli, and G. A. Sawatzky,Two gaps make a high-temperature superconductor?, Rep. Prog. Phys., 71, 062501 (2008).
- [11]N. Miyakawa, J. F. Zasadzinski, L. Ozyüzer, P. Gupta-sarma, D. G. Hinks, C. Kendziora, and K. E. Gray, Predominantly superconducting origin of large energy gaps in underdoped Bi₂Sr₂CaCu₂O₈₊δ from tunneling spectroscopy, Phys. Rev. Lett. 83, 1018 (1999).
- [12]S. H. Pan, E. W. Hudson, A. K. Gupta, K.-W. Ng, H. Eisaki, S. Uchida, and J. C. Davis, STM Studies of the Electronic Structure of Vortex Cores in Bi₂Sr₂CaCu₂O₈₊δ, Phys. Rev. Lett. 85, 1536 (2000).
- [13]Kugler, M., 2000, Ph.D. thesis, University of Geneva.
- [14]Yves Noat Alain Mauger, William Sacks, Statistics of the cuprate pairing states on a square lattice, Modelling Simul. Mater. Sci. Eng. 31, 075010 (2023).arxiv.org/abs/2309.04847
- [15]William Sacks, Alain Mauger and Yves Noat, Universal spectral signatures in pnictides and cuprates : the role of quasiparticle-pair coupling, J. Phys. : Condens. Matter 29, 445601 (2017).DOI
- [16]W. Sacks, T. Cren, D. Roditchev, and B. Douçot, Quasiparticle spectrum of the cuprate Bi₂Sr₂CaCu₂O₈₊δ : Possible connection to the phase diagram, Phys. Rev. B 74, 174517(2006).DOI
- [17]W. Sacks, A. Mauger, Y. Noat, Pair–pair interactions as a mechanism for high-Tc superconductivity, Superconduct. Sci. Technol., 28 105014 (2015).
- [18]W. Sacks, A. Mauger and Y. Noat, Cooper pairs without glue in high-Tc superconductors : A universal phase diagram, Euro. Phys. Lett 119, 17001 (2017).
- [19]Yves Noat, Alain Mauger, William Sacks. Superconductivity in cuprates governed by topological constraints. Physics Letters A 444, 128227 (2022).DOI
- [20]William Sacks, Alain Mauger and Y. Noat, Origin of the Fermi arcs in cuprates : a dual role of quasiparticle and pair excitations, Journal of Physics : Condensed Matter, 30, 475703 (2018).DOI
- [21][21*] R. J. Birgeneau, D. R. Gabbe, H. P. Jenssen, M. A. Kastner, P. J. Picone, T. R. Thurston, G. Shirane, Y. Endoh, M. Sato, K. Yamada, Y. Hidaka, M. Oda, Y. Enomoto, M. Suzuki, and T. Murakami, Antiferromagnetic spin correlations in insulating, metallic, and superconduct- ting La₂−xSrₓCuO₄ Phys. Rev. B 38, 6614 (1988).
- [22]Krusin-Elbaum, T. Shibauchi, G. Blatter, C.H. Mielke, M. Li, M.P. Maley, P.H. Kes, Pseudogap state in overdoped Bi₂Sr₂CaCu₂O₈+y, Physica C 387, Pages 169-174 (2003).DOI
- [23]Maggio-Aprile, I., C. Renner, A. Erb, E. Walker, and Ø. Fischer, Direct vortex lattice imaging and tunneling spectroscopy of flux lines on YBa₂Cu₃O₇−δ, Phys. Rev. Lett. 75, 2754 (1995).
- [24]Neel Haldolaarachchige, S. K. Kushwaha, Quinn Gibson and R. J. Cava, Superconducting properties of BaBi₃, Supercond. Sci. Technol. 27 105001 (2014).
- [25]Adel Nader and Pierre Monceau, Critical field of 2H-NbSe₂ down to 50mK, SpringerPlus, 3, 16 (2014).DOI
- [26]N. Toyota, H. Nakatsuji, K. Noto, A. Hoshi, N. Kobayashi, and Y. Muto and Y. Onodera, Temperature and angular dependences of upper critical fields for the layer structure superconductor 2H-NbSe₂, Journal of Low Temperature Physics, 25, Nos. 3/4, 485 (1976).DOI
- [27]Bartłomiej Andrzejewski, Tomasz Klimczuk, Robert J. Cava, The upper critical field in doped MgCNi₃, Physica C, Volumes 460-462, 706 (2007).
- [28]M. Lubell, Empirical scaling formulas for critical current and critical field for commercial NbTi, IEEE Transactions on Magnetics 19, 754 (1983).DOI
- [29]D.B. Montgomery, H. Wizgall, Measurement of the upper critical field vanadium - gallium alloys, Physics Letters 22, 48 (1966).DOI
- [30]T. P. Orlando, E. J. McNiff, Jr., S. Foner, and M. R. Beasley, Critical fields, Pauli paramagnetic limiting, and material parameters of Nb₃Sn and V₃Si, Phys. Rev. B 19, 4545 (1979).
- [31]M. N. Khlopkin, Anisotropy of the upper critical field and specific heat in V₃Si - a superconductor with cubic symmetry, Condensed Matter 69, 26 (1999).DOI
- [32]S. Foner, E.J. McNiff Jr., B.T. Matthias, T.H. Geballe, R.H. Willens, E. Corenzwit, Upper critical fields of high-temperature superconducting Nb₁−y(Al₁−xGeₓ)y and Nb₃Al : Measurements of Hc₂ > 400 kG at 4.2°K, Phys. Letters, A, 31, 349 (1970).DOI
- [33]Shaoyan Chu and Michael E. McHenry, Synthesis and superconducting properties of a Rb₃C₆₀ single crystal, Phys. Rev. B 55, 11722 (1997).