lampman@heurikon.UUCP (Ray Lampman) (04/26/88)
_________________________________________________________________________ I'm looking for a software package which will be able to help project the physical properties for elemental and multi-atomic structures. _________________________________________________________________________ Most importantly the package must accurately predict weather an atomic arrangement is stable or unstable. It should be able to propose theoretical phase diagrams for elemental and multi-atomic structures. The package should also be able to determine the basic physical properties of elemental and multi-atomic structures, properties like density, hardness, conductivity, malleability, and color. It should accurately predict the reflective properties of metals. And it would be nice if it could predict the transparent properties of glass, plastics, and crystals. _________________________________________________________________________ Now, is this fantasy, or what? _________________________________________________________________________ I have read my own request and am torn between laughter and seriousness. I believe the quantum theory is capable of modeling chemical bonds. But do we have a theory which models the interactions between radiation and matter? If we have sufficient theory, do we have the computing power to complete a valid simulation? Are we missing any parts of this puzzle? I would appreciate any discussion or pointers you can provide. Thank you, -- - Ray Lampman (lampman@heurikon.UUCP)
doug-merritt@cup.portal.com (04/28/88)
Ray Lampman writes: >I'm looking for a software package which will be able to help project the >physical properties for elemental and multi-atomic structures. [...] >I have read my own request and am torn between laughter and seriousness. >I believe the quantum theory is capable of modeling chemical bonds. >But do we have a theory which models the interactions between radiation >and matter? If we have sufficient theory, do we have the computing power >to complete a valid simulation? Are we missing any parts of this puzzle? Your doubts are very well founded. Yes, the theory is (probably) adequate. No, we don't have anything even *close* to the computing power necessary. Simple simulations of single and dual hydrogen atom systems have been carried out with extreme difficulty. Simulations of anything more complex, like say simple sugars, alcohol, methane, etc are totally out of the question. Particle physics has a similar problem with computational complexity of Feynman diagrams (see for instance the very readable QED by Feynman). In general, chemistry is an empirical field *aided* by narrow applications of quantum mechanics. There is currently no hope of predicting general chemical properties wholly from first principles. I talked this over with a friend who's doing research in these areas, and he says that all they can do is use empirical approximations, and keep looking for new theories. That's not to say he implied anyone thought that a theory would be invented that allowed total predictions, though (quite the contrary). On yet another related topic, it turns out that weather prediction is a far more mature field than most people think. The theory works fine, given limitations on the data collection process (precision of measurements, and the resolution of the collecting grid). The principle reason why forecasts are inaccurate beyond a few days out is again because of computational complexity. It's a nonlinear dynamical system which cannot be solved in closed form; it must be simulated one step at a time, and initial inaccuracies of one part in a trillion turn into 100% errors very quickly. See Chaos by James Gleick (also very readable). Back to your original request, depending on your purpose, you might do just as well with a package that is based on empirical data and take it from there. There's lots of molecular modeling packages available for various purposes. I'm not particularly familiar with the details. There was a Scientific American article on the subject sometime in '86. Doug Merritt ucbvax!sun.com!cup.portal.com!doug-merritt or ucbvax!eris!doug (doug@eris.berkeley.edu) or ucbvax!unisoft!certes!doug
colvin@mozart.llnl.gov (Mike Colvin) (04/29/88)
In article <4864@cup.portal.com> doug-merritt@cup.portal.com writes: > >In general, chemistry is an empirical field *aided* by narrow applications >of quantum mechanics. There is currently no hope of predicting general >chemical properties wholly from first principles. Actually this is not entirely correct. There is now a growing field of chemistry known as ab initio quantum chemistry that involves the numerical solution of the molecular Schrodinger equation to calculate chemical properties. Of course these methods yield only approximate solutions, but currently it is possible to predict molecular structures to within about %1 (bond lengths and angles) and energy differences to comparable accuracy. It is possible to use these methods to calculate almost any well-defined chemical property, including dipole moments, vibrational frequencies, etc. Note that although the methods are approximate, they are entirely from first principles, no empirical data is used. In principle these methods are usable for molecules of arbitrary size, but since even the lowest accuracy method (self-consistant field) has complexity 0(n**4) where n is approx. the # electrons, these methods are limited to molecules with a few dozen electrons. The are a variety of ab initio quantum chemistry packages that have been developed by different research groups around the country. (Big wheels in the field include John Pople at Carnegie Mellon, Henry F. Schaefer at U.C. Berkeley (was my Ph.D. advisor), and Bill Goddard at Cal. Tech. ) Currently, only John Pople is freely distributing his package (called Gaussian 86). It is available for a nominal distribution fee. I recommend calling his research group at Carnegie Mellon if you would like a copy. - Mike Colvin (colvin@mozart.llnl.gov)
matt@oddjob.UChicago.EDU (D 1 4 U 2 C) (04/29/88)
colvin@mozart.llnl.gov.UUCP (Mike Colvin) writes:
) There is now a growing field of chemistry known as ab initio
) quantum chemistry ... although the methods are approximate, they
) are entirely from first principles, no empirical data is used.
Wow! This is amazing! How did they calculate the mass of the
electron? :-)
Matt
chas@gtss.UUCP (Charles Cleveland) (04/29/88)
In article <6567@lll-winken.llnl.gov> colvin@mozart.llnl.gov.UUCP (Mike Colvin) writes: )In article <4864@cup.portal.com> doug-merritt@cup.portal.com writes: )> )>In general, chemistry is an empirical field *aided* by narrow applications )>of quantum mechanics. There is currently no hope of predicting general )>chemical properties wholly from first principles. ) ) Actually this is not entirely correct. There is now a growing field )of chemistry known as ab initio quantum chemistry that involves the )numerical solution of the molecular Schrodinger equation to calculate )chemical properties. Of course these methods yield only approximate Well this is entirely charming :-). Not to mention encouraging. I applaud all the people you mention in the text I have deleted and professionally respect the quality of their work. But totally apart from quantum mechanical questions, we can't even take an arbitrary simple potential (of the order of Lennard-Jonesium, for example) and predict what crystal structure it prefers. Even classically. At zero degrees. Of course given two structures, we can tell which is lowest in energy, but that's a different question. The real problems is to find the structure without restricting the problem to certain preconceived possibilities. And not everything is crystalline. Can we predict quasi-periodic structures? What about the possibility of an amorphous ground state? And what about finite temperatures? The original poster wanted to know about phase diagrams. The mind boggles. I would say the poster you responded to stated the situation as per the application of quantum mechanics to general chemical questions far more accurately than you did, but who the hell cares? The calculation of electronic structure for a given arrangement of nuclei is the easy part of the original poster's pipe dream, although it's hard enough. At least the computational rules are clear, and the approximations more or less well understood. The problem is not merely computational. In the case of finding the structure which minimizes the energy of a set of atoms interacting with some given set of potentials, we don't even know how to express the problem in a way conducive to its general solution. Of course if we did, the problem of finding the global minimum would still be of staggering computational difficulty, simply by virtue of the vast number of degrees of freedom in the calculation. Disclaimer: If there is recent literature which I seem to be unaware of, I would appreciate being made aware of it. Nothing could please me more than to learn that my comments here are dated. -- -Life would be so much easier if we could just look at the source code.- Charles Cleveland Georgia Tech School of Physics Atlanta, GA 30332 UUCP: ...!gatech!gtss!chas INTERNET: chas@ss.physics.gatech.edu
lim@cit-vax.Caltech.Edu (Kian-Tat Lim) (04/29/88)
In article <4864@cup.portal.com> doug-merritt@cup.portal.com writes: >Ray Lampman writes: > >>I believe the quantum theory is capable of modeling chemical bonds. >>But do we have a theory which models the interactions between radiation >>and matter? If we have sufficient theory, do we have the computing power >>to complete a valid simulation? Are we missing any parts of this puzzle? > >No, we don't have anything even *close* to the computing power necessary. >Simple simulations of single and dual hydrogen atom systems have been >carried out with extreme difficulty. Simulations of anything more complex, >like say simple sugars, alcohol, methane, etc are totally out of the >question. > Accurate, ab initio (from first principles) simulations of molecules as large as a couple of dozen (light) atoms are *not* that difficult. The only problem really requiring empirical judgement is where to cut off your basis set of orbital functions. See, for example, the GAUSSIAN series of programs developed by Pople at CMU and Hehre at UC Irvine, or the generalized valence- bond implementations from my adviser, Bill Goddard here at Caltech. These programs can produce structures and energies that in some cases been more accurate than contemporary experimental results. Now, admittedly, these are not *complete* solutions to the relevant QM equations [when they are variational, they don't have the lowest possible energy]; but they are adequate for most uses. I believe that Aron Kupperman here has produced provably correct solutions for the "single and dual hydrogen atom systems" mentioned above, "with extreme difficulty," though I am not as familiar with his work. However, this is a level of detail that is unnecessary in almost all cases. Semi-empirical and empirical (force-field) methods allow treatment of much larger systems (I work with proteins of a hundred or so amino acids), but with less accuracy and more caveats. -- Kian-Tat Lim (ktl@wagvax.caltech.edu, GEnie: K.LIM1)
firth@sei.cmu.edu (Robert Firth) (04/29/88)
In article <14715@oddjob.UChicago.EDU> matt@oddjob.UChicago.EDU (D 1 4 U 2 C) writes: >colvin@mozart.llnl.gov.UUCP (Mike Colvin) writes: >) There is now a growing field of chemistry known as ab initio >) quantum chemistry ... although the methods are approximate, they >) are entirely from first principles, no empirical data is used. > >Wow! This is amazing! How did they calculate the mass of the >electron? :-) > Matt When I did this sort of thing, we didn't bother to calculate it. From "first principles", the mass of the electron is 1. You also assume the mass of the proton is infinity, the gravitational constant zero, and voila! you have the Hartree-Fock self-consistent field equations. What amazed me was not that I could get good answers with a machine about the power of a Mac 512E, but that I could get any sane answers at all! (PS: I still have the programs, in Titan Autocode would you believe)
jwm@stdc.jhuapl.edu (James W. Meritt) (04/29/88)
In article <14715@oddjob.UChicago.EDU> matt@oddjob.UChicago.EDU (D 1 4 U 2 C) writes: >Wow! This is amazing! How did they calculate the mass of the >electron? :-) I bet I could come fairly close just using the size of the electron shell, planc's constant, wave-partical duality, and such... It ain't even my field, but that sounds familiar..... Disclaimer: Individuals have opinions, organizations have policy. Therefore, these opinions are mine and not any organizations! Q.E.D. jwm@aplvax.jhuapl.edu 128.244.65.5 (James W. Meritt)
doug-merritt@cup.portal.com (04/30/88)
Matt facetiously asks how they calculate the mass of the electron. In the book "Information Mechanics", the author (Kantor) does in fact calculate, among other things, the mass of the electron. He uses hybrid methods derived from information theory. I asked about a year ago without any firm response: is Kantor's work now mainstream, discredited, or what? The book is about ten years old. P.S> Mike, thanks for the note about ab initio chemistry...apparently my informant was not as well informed as I had thought. Doug Merritt ucbvax!sun.com!cup.portal.com!doug-merritt or ucbvax!eris!doug (doug@eris.berkeley.edu) or ucbvax!unisoft!certes!doug
colvin@mahler.llnl.gov (Mike Colvin) (04/30/88)
In article <247@gtss.UUCP> chas@gtss.UUCP (Charles Cleveland) writes: > >Well this is entirely charming :-). Not to mention encouraging. I >applaud all the people you mention in the text I have deleted and >professionally respect the quality of their work. But totally apart >from quantum mechanical questions, we can't even take an arbitrary >simple potential (of the order of Lennard-Jonesium, for example) and >predict what crystal structure it prefers. Even classically. At zero >degrees. These points are all well taken with regard to cystal structures as well as fluids and amorphorous solids, but the methods I was referring to are used to study molecules composed of a modest number of atoms (up to a few dozen). Certainly there is a lot of chemistry that involves the study of much larger systems, but there are many cases where quantum mechanical calculations have been indepensible to experimental chemists; not as simple qualitative tools, but in providing accurate chemical properties. These cases include the determination of the ground structure and electronic state of the methylene radical, CH2, (initially incorrectly assigned by Herzberg), the assignment of the infrared spectrum of the hydronium ion (H30+), and the confirmation of the decay channel of glyoxal, and many others. > Of course given two structures, we can tell which is lowest >in energy, but that's a different question. The real problems is to >find the structure without restricting the problem to certain preconceived >possibilities. Perhaps I was overzealous in proclaiming quantum chemistry to be fully ab initio (no, we don't calculate the electron mass, fine structure constant, ... :-) since the molecular structures and electron states we use as starting points are based on experiment, semi-empirical rules (Hund's, etc.), and experience. However, starting from these initial configurations the structures are optimized (see below). It is possible that this approach could lead to a gross error (perhaps the structure of benzene really shouldn't be cyclic), but by and large this approach has been very successful (the results are eventually confirmed by experiment). >The problem is not merely computational. In the case of finding the >structure which minimizes the energy of a set of atoms interacting >with some given set of potentials, we don't even know how to express >the problem in a way conducive to its general solution. Of course if >we did, the problem of finding the global minimum would still be of >staggering computational difficulty, simply by virtue of the vast number >of degrees of freedom in the calculation. The problem of finding the global minimum for a particular molecule is indeed a difficult one. However, such optimizations are routinely carried out and you will find that nearly all published theoretical molecular structures have been optimized to where the energy gradients (dE/dRi Ri=coords of each atom in system) are on the order of 10**-6 a.u. (hartrees/bohr), i.e. very small. The number of degrees of freedom is large (3 * # atoms), but not vast, since the systems routinely studied are limited to only a dozen or so atoms. The methods used for these optimizations are standard multi-dimension minimization methods such as Newton-Raphson using gradients and hessians calculated either analytically or by finite differences. I don't want to oversell the utility of ab initio quantum chemistry, but to indicate the it plays only a minor, qualitative role in modern chemistry is simply incorrect. If you look in the Journal of Chemical Physics or JACS (certainly among the flagship journals of chemistry) you will find *many* articles (10-12 an issue in JCP) where these methods are being used for the quantitative prediction of chemical properties. Thanks for raising some interesting points. --Mike Colvin
gwyn@brl-smoke.ARPA (Doug Gwyn ) (04/30/88)
In article <4949@cup.portal.com> doug-merritt@cup.portal.com writes: >In the book "Information Mechanics", the author (Kantor) does in >fact calculate, among other things, the mass of the electron. He >uses hybrid methods derived from information theory. I asked about >a year ago without any firm response: is Kantor's work now mainstream, >discredited, or what? The book is about ten years old. Unless his computation of the mass of the electron is merely a special case of a more general method for calculating all lepton masses, the book discredits itself. There was some excuse for this kind of thing in Eddington's day, because they didn't know of the vast spectrum of elementary particles. There is no excuse for it in modern times.
doug-merritt@cup.portal.com (05/03/88)
Doug Gwyn comments that unless Kantor's methods for calculating the mass of an electron are generalized to all leptons, then it discredits itself. It was in fact just a small part of a very generic methodology; there was nothing magic about "electron mass". The last time I read it was many years ago, so I'm vague on the details. And I didn't know much particle physics at the time (not that I'm an expert now, either). But the basic idea was one of examining the number of bits per state and fiddling around with information transfer rates. Don't flame it on the basis of what I say, though; it's been a long time. It strikes me as being a reasonable idea in general, if you're a believer in conserving causality by preserving locality, etc. On the other hand, not being familiar with the literature, and never having seen a review of his work, he could be considered to be a crank, for all I know. And he hasn't won any Nobel prizes. Towards the end of the book he has some rather interesting speculations about neutrinos that sounded rather testable. I will say that if he *is* a crank, he's an unusual one, in that he demonstrates an intimate familiarity with the standard literature, and critiques sloppiness in a number of published experiments. Doug Merritt ucbvax!sun.com!cup.portal.com!doug-merritt or ucbvax!eris!doug (doug@eris.berkeley.edu) or ucbvax!unisoft!certes!doug