The SubHeat database

Crystal structures. These have been extracted from the Cambridge Database, choosing only entries unaffected by disorder or other errors, and, in case of multiple determinations, selecting the most recent one, which is usually also the one with the lowest / -factor. Room-temperature determinations have been preferred because any simulation potential should preferably be applicable at room-T conditions low-temperature determinations have been selected when room-temperature ones were unacceptable for other reasons. When neutron diffraction data were available, these were selected for the positions of the hydrogen atoms. 

Sublimation enthalpies. The NIST database [5] has been scanned for each compound. When several AHsubi data were present, which was usually the case for the compounds considered (Section 7.6), the choice was made on the basis of (2a) internal consistency when similar values had been obtained by different workers, the average value was assumed (2b) chemical consistency when data were available for a series of chemically similar compounds (e.g. the carboxylic acids, the chlorobenzenes) care was taken to select values which consistently fitted within the series (2c) reliability from personal judgement of the reliability of the research groups, preference being given to data coming from laboratories with a consistent and well documented experience in the business (2d) 

---

Fig. 8.4. A plot of the calculated lattice energy versus the experimental heat of sublimation for the 201 molecules in the SubHeat database. WqEHT force field only for non-hydrogen bonded structures.

Fig. 12.16. The plot of calculated lattice energies against experimental sublimation enthalpies for 172 organic crystals of the Subheat database (Section 8.3). The least-squares straight line through the data points is y = 0.96x R = 0.79).

---