Physics Today

Materials Science Needs and Is Getting Quantitative Methods Gerd Ceder and Jerzy Bernholc

Citation: Physics Today 53(2), 75 (2000); doi: 10.1063/1.882979 View online: http://dx.doi.org/10.1063/1.882979 View Table of Contents: http://scitation.aip.org/content/aip/magazine/physicstoday/53/2?ver=pdfcov Published by the AIP Publishing

Reuse of AIP Publishing content is subject to the terms at: https://publishing.aip.org/authors/rights-and-permissions. Download to IP: 67.180.152.84 On: Wed, 22 Jun 2016 14:26:45 Further, Rhodes states that implo- results suggests that the recessional LETTERS (continued from page 15) sion, the key development beyond velocities measured by Tifft could be

awrence Cranberg compares the “Los Alamos Primer” phase, is written as integral multiples of H0, attributable to Seth Neddermeyer, so that v = nؒH d , where n is an L Oppenheimer with Fermi. As 0 0 postdoc of the former’s and student and Rhodes quotes John Manley, integer and d0 is a basic unit of dis- of the latter’s, I feel that some thing who was there, as saying that Ned- tance (1 Mpc). This equation is basi- should be said. Certainly Fermi was dermeyer faced “stiff opposition” cally a quantized form of Hubble’s a marvelous model for a physicist, from Oppenheimer and others.2 law, and it implies that galaxies are and I don’t know who could stand I stand by my original letter, but located only at certain distances the comparison. Cranberg blames that letter will have served a higher d = nd0 away from us, at least in Oppie for not being, as Fermi was, purpose than evaluating Oppen- the near universe. Just how far out successful in experiment as well as heimer’s role in the A-bomb project this equation would apply is not theory. But who else was? Einstein? if it focuses attention on the under- clear, but it does hold for our near- Feynman? Schwinger? Von Neu- lying and recurring general ques- est galactic neighbor, M31 (the mann? In this, Fermi was probably tions about the requirements for Andromeda galaxy), which is unique in our century. Cranberg leadership of large-scale scientific- known to be approximately 1 Mpc credits development of the A-bomb to engineering endeavors. And I hope away (corresponding to n = 1 in the President Roosevelt, and its use to that both the letter and this ex- above formula). President Truman, and he takes change will continue to stimulate A quantized Hubble’s law might Oppie to task for not having made constructive discussion of those be masked by other effects farther any technical contributions. requirements—surely a topic worthy out, but it does suggest that the I am not happy that the bomb of further discussion in the pages of Hubble constant may play a role in was developed, and much less so both PHYSICS TODAY and APS News. the large-scale universe similar to that it was used, and I do not admire that played by Planck’s constant in Oppie for having been the director of References the atomic domain—that is, in giv- the project. But I have only heard 1. R. Rhodes, The Making of the Atomic ing rise to structure in the universe. good things about his wartime direc- Bomb, Simon and Schuster, New York (1986), p. 460. References tion of Los Alamos, never any criti- 2. Rhodes, pp. 466–67. 1. For a review of his work on redshift cism. In fact, from all that I have LAWRENCE CRANBERG quantization since 1976, see W. G. read, Oppie was an excellent direc- Austin, Texas Tifft, Astrophys. Space Sci. 227, 25 tor. And before the war, he had been (1995). the outstanding leader and teacher 2. For a review of their work on the prob- of theoretical physics in the US. He Does H Play Role in lem, see B. N. C. Guthrie, W. M. Napi- brought into existence the first Am- 0 er, Astron. Astrophys. 310, 353 (1996). erican school of theoretical physics. Universe Like h Does 3. W. Freedman, Sci. Am., November As a student just after the war, I 1992, p. 54. still studied quantum mechanics in Atomic Domain? MAURICE T. RAIFORD from prewar mimeographed notes of f the many redshift studies that ([email protected]) an Oppenheimer course (the teacher Ohave been done over the years, University of Central Florida of my course was Edward Teller). As one of the most interesting has to be Orlando, Florida a young postdoc at the Institute for that of William Tifft of the Steward Advanced Study in Princeton in Observatory in Arizona. He has been 1948–49, where Oppie was the dir- studying and reporting on redshift Materials Science ector, I—like others interested in data for over two decades now, and Needs and Is Getting field theory and in particle phys- has repeatedly found a bunching of ics—eagerly attended the weekly the data around certain values.1 Quantitative Methods seminars he organized. When interpreted in terms of reces- he following comment is prompt- In short, denying Oppenheimer’s sional velocities in the usual way, Ted by my having read Jerzy leading role in physics, especially in these values are integral multiples Bernholc’s article, “Computational US physics, is hardly correct. of a certain basic value—namely, Materials Science: The Era of Ap- JACK STEINBERGER 72 km/s. Although somewhat con- plied Quantum Mechanics,” in your ([email protected]) troversial initially, these basic re- September 1999 issue (page 30). CERN sults were later confirmed by Bruce Although we must be impressed by Geneva, Switzerland Guthrie and William Napier of the the ingenuity that is often displayed Royal Observatory in the UK.2 in large-scale ab initio simulations, RANBERG REPLIES: I welcome the Furthermore, these results have the road from breaking a solid or C responses of Timothy Karpin, also proved to be very close to the molecule in a simulation to the engi- James Osborn, and Jack Steinberger. latest value reported for the Hub- neering concept of “strength” is a To add useful evidence and analysis ble constant, as announced by the long one, and unlikely to be tra-

to the A-bomb story, though, I too Hubble Space Telescope H0 Key Pro- versed by using simulations only. think it best to cite sources. My ject team: 71 km/s per megaparsec Similarly, other relevant engineering reading of Richard Rhodes, for exam- (see PHYSICS TODAY, August 1999, properties, such as corrosion and ple, is that he attributes the “Los page 19). Here it is useful to note fracture resistance, phase (meta)sta- Alamos Primer” to lectures given by that, in her 1992 survey,3 team bility, microstructure formation, and Robert Serber and compiled by asso- coleader Wendy Freedman gave macroscopic transport, are often a 1 ciate lab director Edward Condon. the most probable value of H0 as complex (and unknown) combination J. Robert Oppenheimer’s role was 73 km/(s Mpc). of microscopic phenomena. evidently to convene the lectures. The closeness of all of the above What is the problem? Due to the

FEBRUARY 2000 PHYSICS TODAY 75 Reuse of AIP Publishing content is subject to the terms at: https://publishing.aip.org/authors/rights-and-permissions. Download to IP: 67.180.152.84 On: Wed, 22 Jun 2016 14:26:45 lack of microscopic information, the the “minimum strength” prediction ful for having it brought to our discipline of materials science and is still quite useful. Recent measure- attention. engineering has historically devel- ments made by Richard Smalley’s3 To the best of our knowledge, the oped as an empirical and nonquanti- and Charles Lieber’s groups show possibility of spontaneous P, CP, and tative one. Now that advances in that carbon nanotubes can sustain T violation in strong interactions is computational quantum mechanics at least a 5% strain—making them attributable to T. D. Lee, as reported have made detailed microscopic the “strongest” material known! in his 1973 paper, “A Theory of Spon- information available, we find our- I must also point out that the taneous T Violation.” 1 In a modern selves searching for quantitative progress made in computer science context, Lee considered an hЈ con- materials theories with which to affected my article in another, much densate, which is equivalent to a integrate them. The true challenge, less desirable way. My overconfident region with a nonzero q angle. For therefore, is to develop theories that computer spelling checker changed example, his hЈ condensate induces will lead to the systematic coarse Alex Zettl’s name to Alex Seattle, a nonzero electric dipole moment graining of microscopic phenomena and I, for one, feel profoundly sorry of the neutron (equation 49) and into macroscopic behavior. The prob- and apologize for failing to read the P- and CP-odd contributions to lem, then, is one of detailed knowl- corrections one more time. hadron–hadron scattering ampli- edge of the phenomena at the inter- References tudes (equation 48). mediate scale, rather than one of In two 1974 papers,2 Lee, and Lee computational quantum mechanics. 1. R. E. Rudd, J. Q. Broughton, Phys. Rev. B 58, R5893 (1998). E. B. Tadmor, and Gian Carlo Wick, discussed how GERD CEDER metastable vacuum states, such as ([email protected]) G. S. Smith, N. Bernstein, E. Kaxiras, Phys. Rev. B 59, 235 (1999). those with ∀hЈ¬ Þ 0, arise in effective Massachusetts Institute of Technology 2. M. Buongiorno Nardelli, B. I. Yakob- hadronic theories, and can form Cambridge, Massachusetts son, J. Bernholc, Phys. Rev. B 57, abnormal states of hadronic matter. R4277 (1998). In their 1985 paper,3 Morley and ERNHOLC REPLIES: Although 3. D. A. Walters et al., Appl. Phys. Lett. Schmidt discussed how P, CP, and T B many problems can be addressed 74, 3803 (1999). violations can arise in heavy-ion col- solely by atomistic simulations, JERZY BERNHOLC lisions from regions in which qÞ0, there are many that cannot be, due ([email protected]) but they did not offer a mechanism either to the length or time scales North Carolina State University by which such regions could be gen- involved. As Gert Ceder points out, Raleigh, North Carolina erated. Although they did propose empirical and nonquantitative mod- a signature—a spin correlation be- els provided and continue to provide tween outgoing protons—it remains important guidance in such cases. More on History of However, one of the main emerging a challenging task to measure this P, CP, T Violation in correlation experimentally. theoretical thrusts consists of multi- 4 scale methods, in which microscopic Strong Interactions In contrast, in our 1998 paper, information from atomistic simula- we propose a detailed dynamical n her story entitled “Going for the mechanism of spontaneous P and tions is combined with continuum Gold: First Collisions at RHIC Are mechanics or Monte Carlo methods I CP violation. In both that paper and Set for December” (PHYSICS TODAY, 5 to obtain the required coarse grain- a subsequent work, we also suggest October 1999, page 20), Gloria how this effect would manifest it- ing. (The first—and too long—draft Lubkin mentions a 1998 proposal by version of my article did contain a self in P- and CP-odd correlations of Dimitri Kharzeev, Robert Pisarski, charged pions in heavy-ion collisions, section on such methods.) and Michel Tytgat that a P- and CP- Although multiscale methods are which are measurable on an event- violating metastable state could be by-event basis. still emerging, good progress is being produced in heavy-ion collisions. I 1 made. Ideally, multiscale calcula- would like to inform your readers References tions would proceed in a manner that the original idea for P, CP, and analogous to the multigrid method, l. T. D. Lee, Phys. Rev. D 8, 1226 (1973); T violations in heavy-ion collisions see especially appendix D. in which the information from the due to the local excitation of the vac- 2. T. D. Lee, Phys. Rep. 9, 143 (1974). coarsened solutions is used to recur- uum into an excited state that has T. D. Lee, G. C. Wick, Phys. Rev. D 9, sively accelerate the progress on the the possibility of being CP violating 2291 (1974). finer scales. Conversely, the fine- was given in a 1985 paper by Peter 3. P. Morley, I, Schmidt, Z. Phys. C 26, scale solutions improve the accuracy D. Morley and myself.1 627 (1985). of the coarsening. However, a num- 4. D. Kharzeev, R. D. Pisarski, M. H. G. ber of methodological aspects still Reference Tytgat, Phys. Rev. Lett. 81, 512 (1998). need to be developed. 1. P. D. Morley, I. Schmidt, Z. Phys. C 26, 5. D. Kharzeev, R. D. Pisarski, Turning to the specific example of 627 (1985). hep-ph/9906401. “strength,” let’s focus briefly on the IVAN SCHMIDT DMITRI KHARZEEV nanotube example discussed in my ([email protected]) ([email protected]) article. Ab initio and classical simu- Technical University “Federico ROBERT PISARSKI lations2 predicted that nanotubes Santa Maria” ([email protected]) would be thermodynamically stable Valparaiso, Chile Brookhaven National Laboratory at strains up to 5–6%, and kinetical- Upton, New York ly metastable at significantly greater harzeev, Pisarski, and Tytgat MICHEL TYTGAT strains. Although it has not yet been K reply: We were previously un- ([email protected]) possible to simulate the entire frac- aware of the work of Peter Morley Free University of Brussels ture process at realistic timescales, and Ivan Schmidt, and we are grate- Brussels, Belgium ᭿

76 FEBRUARY 2000 PHYSICS TODAY Reuse of AIP Publishing content is subject to the terms at: https://publishing.aip.org/authors/rights-and-permissions. Download to IP: 67.180.152.84 On: Wed, 22 Jun 2016 14:26:45