Density-Functional Thermochemistry. III. The Role of Exact Exchange
The Journal of Chemical Physics, 1993 · paper · Chemistry
Argues that further gains in density-functional thermochemistry require exact-exchange information, and tests a semiempirical functional combining local-spin-density, gradient and exact-exchange terms on atomization energies, ionization potentials, proton affinities and atomic energies. It fits atomization energies with 2.4 kcal/mol average error.
Checked: PDF from gwern.net, fetched 5 Oct 2026. Page numbers are those of that PDF.
0 findings
None that passed a second check.
2 minor (typos and the like; most checked by the judge only)
Wording Table V gives the average absolute total-energy deviation for Eq. (2) as '6.0' mhartree, but the ten Δ values printed in Table IV average 5.7 mhartree (the text says 'an average error of 6 mhartree').
- p. 5 Table IV, Δ column
+0.010
- p. 5 Table V, Total energies row, Eq. (2) column
6.0
- p. 5 Results text after Table VI
with an average error of 6 mhartree
- p. 5 Table IV deviations / text
with an average error of 6 mhartree and a maximum error of 10
Judge's note: From the page 5 image, the Table IV Δ values are −0.004, −0.004, −0.006, +0.010, +0.010, +0.008, +0.008, +0.001, −0.004 and +0.002 hartree. Each one equals Eq.(2) minus Exact from the printed columns. Their absolute values add to 57 mhartree, so the mean is 5.7. Table V prints 6.0, while the maximum of 10.0 in Table VI does match Table IV. Each printed Δ is rounded to 1 mhartree, so the mean of the unrounded deviations could be about 0.5 mhartree away from 5.7. That makes 6.0 possible but not reproducible from the published values. The text's '6 mhartree' fits either figure. I treat this as a minor consistency point, not a definite error.
Second check: another model agreed
Report a problem with this finding
Typo Section II says the Kohn–Sham exchange energy "will be designated E_X from now on", but Eq. (2) and its definitions write it as E_X^exact ("E_X^exact is the exact exchange energy") and use E_X-type symbols for other quantities (E_X^LSDA, ΔE_X^B88).
- p. 3 Section II, first paragraph
This Kohn-Sham exchange energy will be designated Ex from now on
- p. 3 Eq. (2) and following definitions
is the exact exchange energy
Judge's note: Checked page 3: the left column announces the symbol E_X, and the right column (Eq. 2 and the following definitions) uses E_X^exact. The notation is inconsistent, but the meaning is clear from context, so this is cosmetic.
Second check: none (judge only)