From: Dr P. Murray-R. <pm...@ca...> - 2003-01-29 15:52:25
|
On Jan 29 2003, Miguel Howard wrote: > Peter, > > Well, I took a look at the element tables in OpenBabel and RasMol. And > Imust say that I was quite surprised/discouraged/confused to find > significant discrepancies between the two. Both in terms of covalent > bonding radius and vanderWaals radius. > > (I am not a chemist ... so I just assumed that the numbers would be pretty > close. I recall that you said that some people would want to use their own > radius values, but I assumed we were talking about small variations. :^) Welcome to the wonderful world of chemistry! It is a wonderful synthesis between the exact (e.g. thermodynamic equations) and the classification of experimental observations. Covalent radii are a useful attempt to predict the distance between singly bonded atoms (i.e. X-Y, not X=Y) by assuming that each has an invariant radius. It works very well for first row elements (B, C, N, O F) and not too bad for the next row. It is very variable for metals. In (say) MX4 tetrahedrs we find that the *sum* of the bond orders is fairly constant but that the indidividual bonds are distorted in crystal fields. For Cu(II) we often find 4 short and two long bonds (Jahn-teller distortion). van der Waals - the same thing for non-bonded atoms is even more difficult as it is hard to know when they are in contact. > > I formatted the table below so that you could take a look. Wonderful! > > I still have not looked for implementation of the "bonding algorithm" in > OpenBabel. > > Any comments? > > Miguel > > ----------- > > covalent Radius vanderWaals Radius > # Sym Babel RasMol ratio Babel RasMol ratio > 0 Xx 0 0.720 0.000 0 1.440 0.000 > 1 H 0.23 0.320 0.719 1.2 1.100 1.091 > 2 He 0.93 1.600 0.581 1.4 2.200 0.636 > 3 Li 0.68 0.680 1.000 1.82 1.220 1.492 > 4 Be 0.35 0.352 0.994 1.7 0.628 2.707 > 5 B 0.83 0.832 0.998 2.08 1.548 1.344 > 6 C 0.68 0.720 0.944 1.95 1.548 1.260 > 7 N 0.68 0.680 1.000 1.85 1.400 1.321 > 8 O 0.68 0.680 1.000 1.7 1.348 1.261 > 9 F 0.64 0.640 1.000 1.73 1.300 1.331 I am surprised at several of these values - esp H. I use 0.77 for C and 0.72 for N. I suspect you can forget the vdw He does not really have a covalent radius! > > 10 Ne 1.12 1.120 1.000 1.54 2.020 0.762 > 11 Na 0.97 0.972 0.998 2.27 2.200 1.032 Ne, Na does not really form covalent compounds. I doubt these come from systematic study. > 12 Mg 1.1 1.100 1.000 1.73 1.500 1.153 > 13 Al 1.35 1.352 0.999 2.05 1.500 1.367 > 14 Si 1.2 1.200 1.000 2.1 2.200 0.955 > 15 P 0.75 1.036 0.724 2.08 1.880 1.106 > 16 S 1.02 1.020 1.000 2 1.808 1.106 > 17 Cl 0.99 1.000 0.990 1.97 1.748 1.127 > 18 Ar 1.57 1.568 1.001 1.88 2.768 0.679 > 19 K 1.33 1.328 1.002 2.75 2.388 1.152 > The P value for babel is surprising. > 20 Ca 0.99 0.992 0.998 1.973 1.948 1.013 Ca Does not really form covalent compounds Rest omitted... P. |