Reference Label	Details
Berkowitz 1971	J. Berkowitz, W. A. Chupka, P. M. Guyon, J. H. Holloway, and R. Spohr, J. Chem. Phys. 54, 5165-5180 (1971) Photoionization Mass Spectrometric Study of F2, HF, and DF
Berkowitz 1973	J. Berkowitz and A. C. Wahl, Adv. Fluorine Chem. 7, 147-174 (1973) The Dissociation Energy of Fluorine
Berkowitz 1971a	J. Berkowitz, W. A. Chupka, P. M. Guyon, J. Holloway, and R. Spohr, Adv. Mass Spectrom. 5, 112-115 (1971) Photo-ionization Studies of F2, HF, DF, and the Xenon Fluorides
note F2	Berkowitz 1971 quote 19.008 eV for AE(F+/F2) at 0 K, repeated in Berkowitz 1971a and Berkowitz 1973. They also observe IPF(F2). Berkowitz 1971a, in the discussion following the actual paper, quote the "linearly extrapolated" IPF threshold of 796.3 A, and point out that the difference between AE(F+/F2) - IPF(F2) = 3.43 eV, approx. equal to best then available EA(F) of 3.45 eV. However, the IPF threshold is best interpreted in a manner similar to parent ionization. In particular, in HF the known ion-pair formation threshold seems to correspond to the mid-rise of the first autoionizaing peak, i.e. higher than the mid-rise of the assumed underlying step (probably it corresponds to the turn-over point of the underlying step). A similar position in F2 would correspond to approx. 794.0 +- 0.3 A = 15.615 +- 0.006 eV, or perhaps more conservatively 15.62 +- 0.01 eV. For AE(F+/F2), Berkowitz 1971 report a linearly extrapolated threshold of 652.5 A = 19.001 eV at 80 K, which, when corrected for the internal; energy of 0.007 eV, becomes 19.008 eV. The onset AE(F+/F2) onset sits on top of a broad feature identified as a higher energy re-appearance of the ion-pair process (verified by separately measuring F-). The threshold figure is at a very large expansion with sparse points, which makes it difficult to redo the extrapolation. On the other hand, the overall spectrum is quite compressed. However, from Dibeler 1969, it seems that the threshold has a characteristic curvature toward higher energy, signifying that a linear extrapolation is likely to produce a threshold that is slightly too low. The highest energy point in the expanded figure is at about 649.0 A = 19.104 eV, which is certainly higher than the threshold. The next point is at about 654.0 = 18.958 eV, which is certainly well in the tail. This coarsly brackets the threshold as 651.5 +- 2.5 A. From the overall spectrum, linear extrapolation indeed leads to 652.5 A as a threshold. Assuming some curvature would extrapolate to approx. 625.2 A = 19.010 eV at 80 K or 19.017 eV at 0 K. An uncertainty of +- 0.5 A is equivalent to 0.016 eV.Dibeler 1969 quotes 661 +- 1 A = 18.76 +- 0.03 eV, but that is really the first onset in the tail. An approximate graphical extrapolation of their data suggests 652-653 +- 2 A = 19.00 +- 0.05 eV at 298 K, which, when corrected for the internal energy of 0.027 eV becomes 19.03 +- 0.05 eV.