It was obtained for ORGANIC system (thin film) undergoing photoinduced charge transfer/ separation. Could it be assigned to the formed coupled pair? Any literature data?
Sorry, i don't know the literature in that part of EPR.
Deviations from g=2.0023 are usually due to at least three things:
1) occurrence of effective g-values (usually in transition metals, unlikely in this case)
2) Spin-orbit coupling for S=1/2 (or higher). SOC coupling constant is a small contribution for n=2 (CNOF) level, but if an unpaired electron is localized (or resonance stabilized) on Phosphorous or Sulfur (or something with higher Z) such a radical (S=1/2) could possibly deviate from g=2..
3) Zero-field splitting (D&E) for S>1/2 systems. An S>1/2 could imply the existance of to S=1/2 systems that are coupled, likely what you refer to as a "coupled pair". There should be a symmetric peak.though.
Are you sure with DPPH (or some radical standard) that it really is deviated from g=2? (i do not mean to be insulting in any way, but sometimes the easiest explanation is that the instrument was doing something odd).
[I]If there is some transition metal ion, g can be more than2due to presence of SOC.
[II] It is generally believed that with no SOC in free radicals , the sole contribution will result from the spin motion and like the free electron, g should be=2.0 or 2.0023(relativistic correction included).No doubt, “In most of organic free radicals, the orbital angular moment is quenched. However, in certain radicals (Cl, Br, I, N3, NCS, OH) an orbital degeneracy exists and hence unquenched electronic angular moment (L) is preserved. In such cases, the values of g deviates strongly from g==2.0023”.
[Chem. Rev., 1993, 93,537-70].
[III] In case of CT (photo induction, CTML, CTLM, CTTS), there can be presence of
two charge centers (analogous to two unpaired electrons; one being present at each place) to give two g values (g1,g2) but only the average g=g1+g2/2 be observed at 2.2.
[IV] Lastly, I may humbly add that in such a case, ZFS does exist but alone may not cause such a large increase in g (appxo.10%).