Why Compressed Metal Hydrides are Near-room-temperature Superconductors
Warren E. Pickett · 2025
Read Why Compressed Metal Hydrides are Near-room-temperature Superconductors by Warren E. Pickett (2025) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2508-18429.txt.
Introduction
Warren Pickett asks why some hydrides superconduct so well in twenty twenty-five. Squeezed metal hydrides can superconduct near room temperature. The community picture says a candidate's stability and coupling can be computed from first principles. Those calculations give large electron-phonon coupling, mostly from high-frequency hydrogen vibrations. Predicted critical temperatures match measured ones in known cases.
That much was foreseen by Ashcroft two decades ago. Yet Pickett argues the deeper why is still not understood. His paper focuses on the sharper question of why hydrogen, of all elements, plays this role. He claims the full answer is missing but maps a pathway toward it. Readers get a candid stock-taking of a hot field's foundations.
Catalog record
| CREATOR | Warren E. Pickett |
|---|---|
| YEAR | 2025 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 39 KB |
| VERSION | 1.0 |
| LICENSE | http://creativecommons.org/licenses/by/4.0/ |
| RIGHTS | CC BY |
| SOURCE | arXiv |
Source: https://arxiv.org/abs/2508-18429
Browse the full library · About the Tower of Babel
Full text
The opening of the work, from the library's hosted text:
Why Compressed Metal Hydrides are Near-room-temperature Superconductors Warren E. Pickett Abstract This contribution provides a partial response to the titular statement since, it will be claimed, the “why” is not yet understood, but there is a pathway for achieving a more complete understanding. The sense of the community has been that, given a prospective metal hydride and pressure, the energy landscape can be surveyed computationally for thermodynamic and dynamic stability, the Eliashberg spectral function with its required input (energy bands, phonon modes, coupling matrix elements) can be calculated, and the critical temperature T c obtained. Satisfyingly large values of the electron-phonon coupling strength λ \lambda =2-3 at high mean frequency are obtained, giving very reasonable agreement with existing high T c hydrides. Typically 80-85% of λ \lambda is attributable to high frequency H vibrations. This much was envisioned by Ashcroft two decades ago, so why should there be any angst? This paper addresses more specifically the question why hydrogen? Light mass is indeed a factor, but with possibilities not yet explored. This paper provides a concise overview of related formal developments occurring sporadically over several decades that, when implemented, could resolve the question of why hydrogen, why so high T c . The dearth of success of numerous high throughput searches proposing higher T c materials, especially hydrides, is touched on briefly. Based on as yet unapplied developments in simplifying effects of atomic displacement, it is proposed that there is a straightforward path toward a deeper understanding of “metallic hydrogen superconductivity” in conjunction with added computational efficiency, and that some human-learning should assist in focusing the search for higher T c superconductors. I Introduction The 2014-2015 discovery by Emerets’ group Drozdov2014 ; Drozdov2015 ; Einaga2016 of superconductivity (SC) up to T c =200 K in SH 3 in the 160-200 GPa range ignited a new era in SC research, and especially on understanding its microscopic origin and working toward even higher T c SCs. Remarkably, this compound had been predicted (independently and published earlier) in 2014 by Duan et al. Duan2014 using standard density functional theory (DFT) and Eliashberg formalism, to be a 200 K superconductor in the same pressure range. The following year saw the extension of the calculations and explanation of the origin of very high T c by several groups. Ma2014 ; Papa2015 ; errea2015 ; bernstein2015 ; akashi2015 ; jose2016 ; Quan2016 ; errea2016 . This breakthrough was followed by the prediction and discovery of LaH 10 (T c =250-260K around 200 GPa) LaH10theory ; ME2018a ; ME2018b ; MS2019 and soon after YH 9 (243 K at 200 GPa Peng2017 ; Kong2021 ; Snider2021 ) was predicted also before (or independent of) being confirmed by experimentalists, solidifying their origin as conventional, but now astounding, superconductors. Half a dozen or so compressed metal hydrides with T c above 100 K have been discovered since, displaying similarities and some differences. These metal hydrides are phonon-paired superconductors, as shown most dramatically by the exceptionally large isotope effect Drozdov2015 in SH 3 , which is what theory would predict. After decades of development of the formal theory and then implementation into computational packages, the current state of prediction of new superconductors is on firm grounds. WEP2023 In spite of development and application of high-throughput searches (which necessarily require computational simplifications) of binary and ternary hydrides with several predictions of new and possibly better examples, discovery of new examples has slowed. Quite possibly this limited success is due to some lack of crucial insight into the origin of high T c , rather than or in addition to the now common approach of (i) considering the full phase diagram, (ii) selecting by some criteria the best suspects and checking for stability, then (iii) calculation of the Eliashberg spectral function \alpha^{2}F and thereby T c , that could provide focus to the search for better superconductors. This numerically taxing approach is being addressed by various groups, often using machine learning methods based on a few simple targets, with modest success so far. This paper presents the argument that there is one missing piece of the hydride puzzle that should provide human insight, improve computational efficiency, and provide some focus for computational searches. II The conventional approach Conventional SC theory for elements is straightforward. Using density functional theory methods, calculate the electronic band energies \varepsilon_{kn} , determining the Fermi surfaces, and the phonon frequencies \omega_{q\nu} , on meshes through the Brillouin zone. For each phonon chosen, the potential change due to the phonon is calculated, then the electron phonon matrix elements (functions of kn,k^{\prime}n^{\prime};q,\mu ) evaluated to integrate \varepsilon_{k,n} and \varepsilon_{k+q,n^{\prime}} over the Fermi surfaces. One obtains the Eliashberg spectral function \alpha^{2}F(\omega) , from which the electron-phonon coupling (EPC) strength \lambda is obtained for a single element SC as given by McMillan: McMillan1968 \displaystyle\lambda=\int\frac{2}{\omega}\alpha^{2}F(\omega)d\omega\rightarrow\frac{N(0)I^{2}}{M\omega_{2}^{2}}. (1) F(\omega) is the phonon density of states, and \alpha^{2}(\omega) is the squared EPC matrix element averaged over all phonons of frequency \omega . The final expression is exact for elements, in which case the el-ion matrix elements I^{2} hides the complication that arises for compounds; for elements it is \displaystyle I^{2}=\frac{\sum_{k^{\prime}n^{\prime}}\sum_{kn}|<kn|\frac{dV}{d\vec{R}}|k^{\prime}n^{\prime}>|^{2}\delta(\varepsilon_{kn})\delta(\varepsilon_{k^{\prime}n^{\prime}})}{\sum_{k^{\prime}n^{\prime}}\delta(\varepsilon_{k^{\prime}n^{\prime}})\sum_{kn}\delta(\varepsilon_{kn})}, (2) where V(r,\{\vec{R}\}) is the electronic potential for ions at positions \{\vec{R}\} and the derivative is respect to displacement of the atom at \vec{R} , and the double average is over k and k^{\prime} over the Fermi surface. Binary and ternary compounds cause an essential complication. Each phonon involves displacement of all atoms, and the preferred method is to calculate the kernel of the matrix element for each phonon, i.e. the potential change, to evaluate the matrix element GiustinoRMP \displaystyle g_{k,n;k+q,n^{\prime};q\nu}=\langle k,n|\hat{\epsilon}_{q\nu}\cdot\frac{dV}{d\hat{\epsilon}_{q,\nu}}|k+q,n^{\prime}\rangle (3) for states k and k+q on the Fermi surface. Each integrand of the matrix element involves displacements of all atoms involved in phonon q,\nu , then the scalar product of the eigenvector with the gradient of potential, then evaluation of the matrix element. The potential V is the sum of the pseudopotential and the density dependent potential V^{DFT} , to which we return below. The square |g_{k,n;k+q,n^{\prime};q\nu}|^{2} occurs in \alpha^{2}F . There are so many integrands, i.e. matrix elements, that they have not been studied, indeed it is unclear just how they should be analyzed. These matrix elements are accumulated into \alpha^{2}F(\omega) involving a sum over all phonons \{q,\nu\} , and over k and k+q restricted to near the Fermi surface. To begin to get to the point, each phonon displacement (hence, matrix element) involves contributions from every atom in the cell, and each phonon is calculated separately. It should also be noted that quantities are no longer simply dependent on N(0) but on how the densities of states are distributed over the various atomic orbital contributions around the Fermi surfaces. The derivative is with respect to each phonon displacement eigenvector \hat{\epsilon}_{q\nu} , and calculation of the matrix element is widely understood by the computational community as costly. The proliferation of indices, sufficient q -mesh grid (see comments in the Appendix), and number of atoms in the unit cell makes the computational task evident. Decreasing each q -mesh separation by a factor of 2 leads to a corresponding factor of 8 in 3D, and increasing the number of atoms in the unit cell