摘要
Abstract More than 50 crystals containing 7,7,8,8-tetracyanoquinodimethane (TCNQ) in various guises are classified into a limited number of structural types and the possible assignment of charge states by diffraction and spectroscopic methods is compared. In the crystalline state the 7,7,8,8-tetracyanoquinodimethane (TCNQ) moiety is found in mixed stack π–π* molecular compounds with neutral TCNQ as acceptor and a variety of neutral donors, and as five types of salt, three of which contain only TCNQ1− and have either (i) segregated monad TCNQ stacks, (ii) mixed-stack {[cation+][TCNQ1−]} arrangements or (iii) isolated {2[cation+] · [(TCNQ1−)2]} π-dimers, while the remaining two contain both TCNQ0 and TCNQ1− and have compositions (iv) {[cation+]]} and (v) {[cation+]2 · }. Each family is found with a characteristic isostructural packing arrangement, there being differences of detail among the members; there are some exceptions to the overall rule, usually with combinations of standard packing elements forming a non-standard overall arrangement. Each structural group has been analysed in the same way. First, the crystal chemistry is described, then various probes of increasing sensitivity are applied to establish the charge state of the moieties and their relation to the packing arrangement. The several probes are (i) X-ray diffraction, giving moiety geometries (bond lengths and deviations from planarity) (ii) infra-red and resonance Raman spectroscopy for identification of charge states (iii) electron spin resonance and magnetic measurements (where applicable) to give further details of moiety structure. Bond lengths have been surveyed for the various types of TCNQ moiety. The differences are not large, the bond most sensitive to charge being the extra-ring double bond (c), while bond length differences for adjacent single-bond double-bond pairs (intra-ring (b – a) and extra-ring (c – d)) also provide a way of distinguishing different charge states. Although bond lengths have provided, for almost 40 years, a popular way to assess moiety charge, our present review suggests that much caution is required in the application of this method not only to individual structures determined at room temperature but even to those determined at very low temperatures. Resonance Raman spectroscopy provides an alternative method of assessing moiety charge and is determinative with singly charged TCNQ moieties but not with other TCNQ species; both methodology and correlation with diffraction results require further investigation. Most studies of electrical conductivity lack detail about orientation and temperature dependence and thus are difficult to relate to packing arrangements. ESR measurements generally confirm strong antiferromagnetic coupling within π-dimers. 1An abbreviated version was presented at the 2007 Salt Lake City Meeting of the American Crystallographic Association on the occasion of the Fankuchen Award to FHH. Keywords: TCNQ structural familiesX-ray diffractionRaman resonance spectroscopyassessing moiety charge Acknowledgements We are grateful to Professors E. A. Halevi and Shammai Speiser and Dr Aharon Blank (all at Technion) for helpful discussions, and to Professor J. S. Miller (Utah) for his interest. Notes 1An abbreviated version was presented at the 2007 Salt Lake City Meeting of the American Crystallographic Association on the occasion of the Fankuchen Award to FHH. Notes 1. Acker and Hertler remark 'Of particular interest is the ease with which this compound accepts one electron to form stable anion radical derivatives'. 2. We use 'moiety' to include various possible charge states of a molecule. 3. A single polymorph is a contradiction in terms (an oxymoron). 4. Cambridge Structural Database (CSD) Refcodes are included as well as conventional references. 5. The charge distributions for () and () must be determined for each individual salt. 6. Most crystal structure diagrams have been prepared with CrystalMaker software (Users Guide, 6th Edition, David Palmer, 2007). Our diagrams generally differ from those in the literature and thus two different points of view are available. 7. This is a compromise between 'precision of structure determination' and 'sample of adequate size'. 8. (Dbbpy.Pd.dmid) is. (4,4′-di-t-butyl-2,2′-bipyridine)-(1,3-dithiole-2-one-4,5-dithiolato)-Pd(II). The Pt compound with TCNQ is isomorphous. 9. 'range' is defined as (highest value – lowest value). 10. We deliberately employ a non-committal term for this group. 11. Shibaeva and Atovmyan (6) write: 'One of the fundamental questions that arises in the interpretation of the physical properties of complex ion-radical TCNQ salts, in particular the temperature dependent magnetic and electrical properties, is the question of the distribution of anionic charges in the solid state i.e. whether the charge is concentrated on particular TCNQ units or is delocalized. Unfortunately, structural studies give almost no answer to this question'. Thirty years have passed but the situation has not changed much. We add that the distinction between §1 and §2 above was well-appreciated by Shibaeva and Atovmyan. 12. There are two typographical errors; the overbar is missing from the space group which is (#2) and the N … N hydrogen bond distance given as 2.267 Å is 2.967 Å. 13. This unit cell was used in the structure determination; the reduced cell is 6.8347 7.773 14.093 Å, 105.445 93.211 106.431°. 14. CSD assigns P1 (#1), Z = 1 to both. 15. 'Isostructural' is used in the sense that there is an overall resemblance between the two structures; 'isomorphism' would have been used if there had been close similarity. 16. Powder diffraction studies do not appear to have been made.