Hydrogen bonding competition between the polyprotic acid cation [(η⁵-C₅H₄COOH)₂Co]⁺ and the polyprotic acid anion [H₂PO₄]⁻

---

Abstract

The organometallic dicarboxylic acid cation [(η⁵-C₅H₄COOH)₂Co]⁺ crystallises with the dihydrogen phosphate acid anion [H₂PO₄]⁻ from water to form the hydrated crystal [(η⁵-C₅H₄COOH)₂Co]⁺[H₂PO₄]⁻·H₂O (1), and with its zwitterionic form, one anion [H₂PO₄]⁻ and two neutral acid H₃PO₄ molecules to form the anhydrous co-crystal [(η⁵-C₅H₄COOH)₂Co]⁺[H₂PO₄]⁻·[(η⁵-C₅H₄COOH)(η⁵-C₅H₄COO)Co]·2[H₃PO₄] (2), which can also be formulated as 2{[(η⁵-C₅H₄COOH)₂Co]⁺[H₂PO₄]⁻}·[H₃PO₄], because of the competition in the possession of the proton in the O–H⋯O hydrogen bond between the organometallic and inorganic moieties.

---

Introduction

Hydrogen bonding (HB) interactions between ions have been a subject of continuing investigation in solution¹ and in the solid state² as well as being the subject of theoretical studies by both empirical and ab initio methods.³ More recently HB interactions between ions have begun to be utilised systematically in the crystal engineering of new materials with remarkable achievements.^4,5 The reason for this interest stems from the fact that HB between ions combines the strength of the Coulombic field generated by the ions with the directionality and reproducible topology of hydrogen bonding interactions.⁶ It has been pointed out that the contribution of hydrogen bonding interactions to the cohesive energy of a crystal depends on the charge carried by the hydrogen-bridged ions: if the HB joins ions of opposite charge [more generally ⁽⁺⁾X–H⋯Y⁽⁻⁾] then the bonding contribution adds to the favourable (+)⋯(−) Coulombic interaction between donor and acceptor systems; whereas, if the ions carry the same charge, e.g.⁽⁻⁾X–H⋯Y⁽⁻⁾ [but also ⁽⁺⁾X–H⋯Y⁽⁺⁾], then the contribution reduces the repulsive electrostatic terms [i.e. (−)⋯(−) or (+)⋯(+)] arising from the Coulombic interactions between like charges. The possible combinations of ionic charges on the donor–acceptor systems are shown below: note that the only ‘true’ charge-assisted interactions are those of the ⁽⁺⁾X–H⋯Y⁽⁻⁾ and X–H⋯Y⁽⁻⁾ type.

When there is competition for the ‘possession’ of the H-atom, i.e. in the case of proton transfer from acid to base [e.g. X⊕+⊕H–Y⊕→⊕⁽⁺⁾X–H⊕+⊕Y⁽⁻⁾], it is the stronger acid that gives the proton away, whilst the species accepting the proton is the stronger base. This way of thinking is almost reversed when considering hydrogen bonding donation within intermolecular or interionic X–H⋯Y interactions. In this latter situation the HB donor is the electrophile (X–H), while the HB acceptor (Y) acts as the nucleophile. Considerable effort has been made to correlate concepts such as acidity and basicity to HB formation and stability.⁷ Correlations between the pK of the acids and HB distances are available.⁸ It has been recently argued, however, that hydrogen bond ability can be better appreciated by examining the charge distribution in the HB system rather than its acid–base chemistry.⁹

In this paper we discuss a rather unusual competition between the polyprotic organometallic acid cation [(η⁵-C₅H₄COOH)₂Co]⁺ and the polyprotic inorganic acid H₃PO₄. The two acids have been co-crystallised from water. The crystalline products [(η⁵-C₅H₄COOH)₂Co]⁺[H₂PO₄]⁻·H₂O (1) and [(η⁵-C₅H₄COOH)₂Co]⁺[H₂PO₄]⁻·[(η⁵-C₅H₄COOH)(η⁵-C₅H₄COO)Co]·2[H₃PO₄] (2) are essentially ionic in nature, thanks to the cationic and anionic nature of the two acids. The hydrated compound 1 is obtained first, while 2 crystallises only when the solution is concentrated by solvent evaporation. The high quality of the diffraction data allowed unambiguous definition of the H-atom positions within the O–H⋯O bridges.

The supramolecular chemistry of the cobalticinium acid has been extensively studied by us.¹⁰ In this earlier study our interest had been focused on the participation of the organometallic acid in hydrogen bonded networks as a mono-cation when fully protonated, as a zwitterion when mono-deprotonated and as a mono-anion when both protons are removed.

Experimental

Crystal synthesis

The zwitterion [(η⁵-C₅H₄COOH)(η⁵-C₅H₄COO)Co] was prepared according to the synthesis previously reported.¹¹ Phosphoric acid was purchased from Aldrich.

Synthesis of 1 and 2. Yellow powder of the zwitterion [(η⁵-C₅H₄COOH)(η⁵-C₅H₄COO)Co] (105 mg, 0.38 mmol) was dissolved in 25 ml of double-distilled water under stirring at room temperature. Anhydrous phosphoric acid in slight excess (50 mg, 51 mmol) was added to the solution. Slow evaporation of the solvent afforded first crystals of 1 (elongated yellow needles); evaporation of the remaining water caused formation of crystalline 2 (yellow prisms).

Crystal structure characterisation

X-Ray diffraction data collections were carried out on a NONIUS CAD-4 diffractometer for 1 and on a Bruker SMART diffractometer for 2. Crystal data and details of measurements are reported in Table 1. SHELXL97^12a was used for structure solution and refinement based on F². SCHAKAL99^12b was used for the graphical representation of the results. Common to all compounds: Mo-Kα radiation, λ⊕=⊕0.71073 Å; monochromator, graphite; T⊕=⊕293 K. All non-H atoms were refined anisotropically. The positions of the H(CH) hydrogen atoms in 1 and 2 were added in calculated positions and refined riding on the corresponding C atoms. H atoms belonging to the COOH groups, to the phosphate anions and to the phosphoric acid molecules were observed in the Fourier maps and refined, with the exception of the H atoms involved in the O(10)–H–O(7), O(8)–H–O(9) and O(6)–H–O(16) intermolecular interactions in 2 – these atoms are in fact disordered, along the O⋯O vectors, over two positions, which were assigned occupancy factors of 0.5 and not refined. The computer program PLATON^12c was used to analyse the geometry of the hydrogen bonding patterns. In order to evaluate C–H⋯O interactions the C–H distances were normalised to the neutron derived value of 1.08 Å.

Table 1 Crystal data and details of measurements for crystalline 1 and 2^a

Formula	1	2
	C12H13CoO9P	C24H27Co2O20P3
a Click b009265g.txt for full crystallographic data (CCDC 153205–153206).
M	391.12	846.23
System	Monoclinic	Monoclinic
Space group	P21/n	P21/c
Z	4	4
a/Å	6.839(5)	13.4021(7)
b/Å	16.563(5)	11.6199(7)
c/Å	12.661(3)	19.471(1)
α/°	90	90
β/°	97.18(5)	91.465(2)
γ/°	90	90
U/Å^3	1423(1)	3031.3(3)
F(000)	796	1720
μ(Mo-Kα)/mm^−1	1.366	1.345
θ range/°	3–25	2–34
Min.and max.transmission	0.86–1.00	0.77–0.79
Measured reflections	2583	42814
Unique reflections	2468	11 500
Refined parameters	215	439
Goodness of fit on F^2	0.817	0.720
R1 [on F, I⊕>⊕2σ(I)]	0.0298	0.0383
wR2(on F^2, all data)	0.0833	0.0790

---

Results and discussion

Compound 1 contains the organometallic cation [(η⁵-C₅H₄COOH)₂Co]⁺, one [H₂PO₄]⁻ anion and one hydrogen bonded water molecule; it can therefore be formulated as [(η⁵-C₅H₄COOH)₂Co]⁺[H₂PO₄]⁻·H₂O. Fig. 1 shows the supramolecular arrangement of two [(η⁵-C₅H₄COOH)₂Co]⁺ cations and two [H₂PO₄]⁻ anions together with two water molecules. Since all hydrogen atoms could be located from the Fourier maps the discussion of the hydrogen bonding interactions is unambiguous (see Table 2). The following features can be noted: the two organometallic cations are in their fully protonated form and donate hydrogen bridges towards the phosphate oxygen atoms of the central [H₂PO₄]⁻ anions with ⁽⁺⁾O(H)⋯O⁽⁻⁾ distances of 2.513(2) and 2.576(2) Å, respectively; the two [H₂PO₄]⁻ anions form a hydrogen-bridged ring with an ⁽⁻⁾O(H)⋯O⁽⁻⁾ distance of 2.647(2) Å (repeated by centrosymmetry to form the ring). the P–O bond length distribution is in agreement with the observed location of the OH group [1.516(2), 1.496(2), versus 1.564(2), 1.557(2) Å]; the water molecules appear to participate in bifurcated O(H)⋯O⁽⁺⁾ and O(H)⋯O⁽⁻⁾ interactions with the organometallic cation on one hand and with the hydrogen phosphate anion on the other [3.095(2) and 3.063(2) Å]; the water molecules also accept hydrogen bridge donation from the phosphate anions [⁽⁻⁾O(H)⋯O_water distance 2.764(2) Å].

Fig. 1 The supramolecular arrangement of two [(η⁵-C₅H₄COOH)₂Co]⁺ and two [H₂PO₄]⁻ anions together with two water molecules in crystalline 1.

Table 2 Relevant hydrogen bonding interactions in crystalline 1 [C–H distances normalized to the neutron value (1.08 Å); (C)H⋯O⊕<⊕2.60 Å; e.s.d.s⊕=⊕2 for all distances and angles]

There is a remarkable packing feature in crystalline 1 that needs to be addressed at this stage. Fig. 2(a) shows how the acid cations are organized to form a belt surrounding the hydrated [H₂PO₄]⁻ anion dimers. The interactions between cations are based on the formation of pairs of C–H⋯O bridges between the cyclopentadienyl hydrogens and the carboxylic oxygens. The (C)H⋯O separations in 1 are in the range 2.329(2)–2.506(2) Å and are within the values expected for these interactions in organometallic crystals.¹³ The cations are piled up so as to form a kind of continuous channel in the crystal. A space-filling representation is shown in Fig. 2(b). The interest in this packing stems from the surprisingly close analogy between the arrangement in 1 and the organization of the zwitterions [(η⁵-C₅H₄COOH)(η⁵-C₅H₄COOH)Co] in the three-hydrated crystalline material [(η⁵-C₅H₄COOH)(η⁵-C₅H₄COOH)Co]·3H₂O observed previously.¹⁴ A space-filling representation of this latter species is shown in Fig. 2(c) and ought to be compared with the packing shown in Fig. 2(b). The structural analogy is evident also from a comparison of the unit cell parameters [a, b, c, β, U, 6.839(5), 16.563(5), 12.661(3) Å, 97.18(5)°, 1422.9(12) Å³versus 6.707(3), 14.997(4), 13.453(4) Å, 99.84(1)°, 1333(8) Å³ in 1 and in the hydrated zwitterion, respectively].

Fig. 2 (a) The organization of the acid cations to form a belt surrounding the hydrated [H₂PO₄]⁻ anion dimers in 1. (b) A space-filling representation of how the cations are piled up so as to form a continuous channel in the crystal. Click image or 2b.htm to access a 3D representation. Compare this arrangement with the one observed in crystalline [(η⁵-C₅H₄COOH)(η⁵-C₅H₄COO)Co]·3H₂O¹⁴ (c), where the channels inside the hexameric units are filled with water molecules. [H atoms in (b) and (c) are not shown for clarity.] Click image or 2c.htm to access a 3D representation.

Let us now consider compound 2. The system contains five independent moieties (see Fig. 3). On the basis of the distribution of the structural parameters and hydrogen bridges, compound 2 can be formulated either as [(η⁵-C₅H₄COOH)₂Co]⁺[H₂PO₄]⁻·[(η⁵-C₅H₄COOH)(η⁵-C₅H₄COO)Co]·2[H₃PO₄] or as 2{[(η⁵-C₅H₄COOH)₂Co]⁺[H₂PO₄]⁻}·[H₃PO₄]. The former corresponds to an acid/anion pair, viz. [H₂PO₄]⁻ and [(η⁵-C₅H₄COOH)₂Co]⁺ as in 1, which is co-crystallised with a zwitterion molecule [(η⁵-C₅H₄COOH)(η⁵-C₅H₄COO)Co] and two un-deprotonated H₃PO₄ units, whereas the latter corresponds to the situation observed for 1, but where the organometallic moieties are in the protonated cationic form. The reason for this ambiguity arises from the competition for the hydrogen atom possession along the O⋯H⋯O system as discussed in the following.

Fig. 3 Ball-and-stick representation of the asymmetric unit in crystalline 2. Note how the H atoms along the O(6)⋯O(16), O(7)⋯O(19) and O(8)⋯O(9) directions are disordered over two positions.

The complex identified by Co(1) is a fully protonated species, i.e. an organometallic cation as in 1. HB parameters are reported in Table 3. The two COOH groups participate in two and in one hydrogen bonding interactions with the phosphate moieties and the phosphoric acid molecule, respectively. There are two O⋯O separations [O(3)⋯O(18) 2.556(2), and O(2)⋯O(17) 2.530(2) Å] that suggest an interaction between a protonated COOH group and a hydrogen phosphate anion, i.e. a ⁽⁺⁾O(H)⋯O⁽⁻⁾ hydrogen bonding interaction. The HB donors are easily identified on the basis of the hydrogen atom locations from Fourier maps and from the length of the C–O(H) bonds [C(6)–O(2) 1.304(2), and C(12)–O(3) 1.316(2) Å] within the COOH groups.

Table 3 Relevant hydrogen bonding interactions in crystalline 2 [C–H distances normalized to the neutron value (1.08 Å); (C)H⋯O⊕<⊕2.60 Å; e.s.d.s⊕=⊕2 for all distances and angles]

The identification of complex [Co(2)] as a zwitterion or a cation is more controversial. As detailed in the Experimental section, two distinct hydrogen atom positions were observed along the O(6)⋯O(16), O(7)⋯O(10) and O(8)⋯O(9) vectors, corresponding to a double well potential, as observed for moderate to strong hydrogen bonds.¹⁵ The short value observed for the O(6)⋯O(16) distance [O⋯O 2.467(2) Å] is in agreement with the presence of a hydrogen bond of the O–H⋯O⁽⁻⁾ type,¹⁶ which can either be described as COO⁽⁻⁾⋯HO(P) or COOH⋯⁽⁻⁾O(P), in agreement also with the distribution of C–O and P–O distances. The disorder observed for the hydrogen atoms in the O(7)⋯O(10) and O(8)⋯O(9) systems [O⋯O 2.537(2) and 2.513(2) Å, respectively], on the other hand, does not alter the charge distribution between the organometallic and inorganic moieties, and can be rationalized by a two-fold disorder of the hydrogen-bonded ring, which is also reflected in the intramolecular C–O and P–O distances.

The piles formed by the zwitterion and the cation are shown in Fig. 4. There is, once again, a marked structural analogy between 2 and compound 1, hence also with [(η⁵-C₅H₄COO)(η⁵-C₅H₄COOH)Co^I]·3H₂O¹⁴ [compare Fig. 4 with Figs. 2(b) and (c)]. The formation of hexamers around the anionic or solvent piles is quite intriguing. The organometallic [(η⁵-C₅H₄COOH)₂Co]⁺ system appears to invariably prefer interactions with alternative acceptor systems (solvent molecules or acid anions) when available. The HB 2-D network formed by the phosphate anions and the phosphoric acid molecules is shown in Fig. 4(b).

Fig. 4 (a) Space-filling representation of the hexameric units, formed by the neutral and cationic organometallic moieties, surrounding an anionic pile in crystalline 2. Click image or 4a.htm to access a 3D representation. (b) Ball-and-stick (top) and space-filling (bottom) representation of the 2D network formed by phosphoric acid units and by the disordered hydrogen phosphate/phosphoric acid unit (in yellow) in 2. Click image or 4b.htm to access a 3D representation.

The hydrogen bonding structural parameters grouped in Tables 2 and 3 indicate a fairly precise ‘hierarchy’ in the distance parameters. It has been pointed out¹⁶ (and argued against¹⁷) that the length of the HB interactions reflects only the strength of the local interaction and has no implications on the overall stability of the aggregate. In the case of ionic crystal cohesion it will, in fact, depend upon the overall balance of repulsive and attractive interactions, of a Coulombic nature or arising from other non-covalent interactions such as hydrogen bonds.¹⁸

While O–H⋯O interactions concerning carboxylic groups have been extensively studied¹⁹ not much is known on the average parameters and distribution of O–H⋯O interactions involving the dihydrogen phosphate anions. Fig. 5 shows a histogram of the distribution of ⁽⁻⁾O(H)⋯O⁽⁻⁾ distances extracted from the CSD²⁰ for all structures containing the [H₂PO₄]⁻⋯[H₂PO₄]⁻ systems. On a total of 67 observations, the mean value O⋯O is 2.580₄ Å, whereas the lowest 10% quantile (lowest decile), which is indicative of the spread of the distribution and of the relevance of short interactions, occurs at 2.537 Å. There is only one very short O⋯O separation (2.460 Å) and this corresponds to the structure of HISTPA10²¹ (L-histidinium dihydrogen orthophosphate co-crystallised with orthophosphoric acid) which contains chains of anions interacting both with the counterion and with the acid molecules. The data in Tables 2 and 3 show that the ⁽⁻⁾O(H)⋯O⁽⁻⁾ hydrogen bonding parameters in compounds 1 and 2 fall within the statistical distribution of the interactions shown in Fig. 5.

Fig. 5 The distribution of ⁽⁻⁾O(H)⋯O⁽⁻⁾ distances extracted from the CSD, for a total of 67 observations for ⁽⁻⁾O(H)⋯O⁽⁻⁾ involving the [H₂PO₄]⁻⋯[H₂PO₄]⁻ systems. The mean value is 2.580₄ Å, whereas the lowest decile occurs at 2.537 Å.

It is worth comparing the average values obtained for ⁽⁻⁾O(H)⋯O⁽⁻⁾ HB involving the [H₂PO₄]⁻⋯[H₂PO₄]⁻ systems with the distribution obtained for the interactions involving polycarboxylic acid anions.^16a The O⋯O distance in this latter sample is shorter, both in average (2.528(5) Å) and in lowest decile value (2.462 Å), than in the case of dihydrogen phosphate anions. Whether the difference between the two samples arises from an effectively stronger local interaction between the COOH and the COO⁽⁻⁾ groups, with respect to that between the POH and the PO⁽⁻⁾ groups, or to a higher contribution from the repulsive (−)⋯(−) term between the small [H₂PO₄]⁻ anions with respect to polycarboxylic acid molecules requires further study, and in the near future will possibly be tackled by theoretical means.

Conclusions

The neutral organometallic zwitterion [(η⁵-C₅H₄COOH)(η⁵-C₅H₄COO)Co] is protonated by the inorganic acid H₃PO₄ to form the dicarboxylic acid cation [(η⁵-C₅H₄COOH)₂Co]⁺. The two acids compete in the possession of the hydrogen atoms within the hydrogen bridges. In the hydrated compound 1 the organometallic moiety is fully protonated, whereas in the anhydrous compound 2 the proton involved in the shortest OH⋯O bond ‘hesitates’ between the organometallic and the inorganic units. Compound 2 can thus be formulated as both [(η⁵-C₅H₄COOH)₂Co]⁺[H₂PO₄]⁻·[(η⁵-C₅H₄COOH)(η⁵-C₅H₄COO)Co]·2[H₃PO₄] and as 2{[(η⁵-C₅H₄COOH)₂Co]⁺[H₂PO₄]⁻}·[H₃PO₄]. This indicates that – at least in the solid state – the two acids have comparable acidity and hydrogen bonding donor–acceptor capacity.

The structural parameters concerning the inter-anionic HB of the ⁽⁻⁾O(H)⋯O⁽⁻⁾ type involving the [H₂PO₄]⁻⋯[H₂PO₄]⁻ chain systems have been investigated via CSD analysis and compared with the data available for ⁽⁻⁾O(H)⋯O⁽⁻⁾ interactions between carboxylic groups. It has been shown that these latter interactions are, on average, shorter than those involving dihydrogen phosphate anions.

We have shown that hybrid organometallic–inorganic crystalline systems can be constructed by utilizing inorganic acids and organometallic acids in ionic systems. Inter-ionic hydrogen bridges are a powerful means to exploit the directionality (and hence the predictability and reproducibility) of the hydrogen bonding interaction in the construction of ionic crystals. This may have useful implications in crystal engineering.²² Furthermore, dihydrogen phosphate salts are known to be useful materials in NLO applications, and the utilization of organometallic systems opens up new routes towards the preparation of systems exhibiting second harmonic generation.²³

Acknowledgements

Financial support from the Universities of Sassari and of Bologna (project Innovative Materials) and from MURST (project Solid Supermolecules) is acknowledged.

References

1. See for example: (a) J. Emsley, Chem. Soc. Rev., 1980, 9, 91; (b) M. Meot-Ner (Mautner), J. Am. Chem. Soc., 1984, 106, 1257; (c) M. Meot-Ner (Mautner) and L. W. Sieck, J. Am. Chem. Soc., 1986, 108, 7525; (d) A. Novak, Struct. Bonding, 1973, 177.
2. (a) D. Wiechert and D. Mootz, Angew. Chem., Int. Ed., 1999, 38, 1974; (b) D. Mootz and M. Shilling, J. Am. Chem. Soc., 1992, 114, 7435; (c) D. Braga and F. Grepioni, J. Chem. Soc., Dalton Trans., 1999, 1; (d) D. Braga and F. Grepioni, Coord. Chem. Rev., 1999, 183, 19.
3. (a) H. Umeyama and K. Morokuma, J. Am. Chem. Soc., 1977, 99, 1316; (b) J. Emsley, O. P. A. Hoyte and R. E. Overill, J. Am. Chem. Soc., 1978, 100, 3303; (c) V. Bertolasi, P. Gilli, V. Ferretti and G. Gilli, Chem. Eur. J., 1996, 8, 925; (d) L. Turi and J. J. Dannenberg, J. Am. Chem. Soc., 1993, 97, 12197; (e) L. Turi and J. J. Dannenberg, J. Am. Chem. Soc., 1994, 116, 8714.
4. (a) M. M. Matsushita, A. Izuoka, T. Sugawara, T. Kobayashi, N. Wada, N. Takeda and M. Ishikawa, J. Am. Chem. Soc., 1997, 119, 4369; (b) See also: J. Chem. Soc., Dalton Trans., 2000, pp. 3705–3998 (whole issue: Proceedings of Dalton Discussion No.3 – Inorganic Crystal Engineering)..
5. (a) D. Braga and F. Grepioni, Chem. Commun., 1996, 571; (b) M. W. Hosseini and A. De Cian, Chem. Commun., 1998, 727; (c) L. Brammer, D. Zhao, F. T. Ladipo and J. Braddock-Wilking, Acta Crystallogr., Sect. B, 1995, 51, 632; (d) C. B. Aakeröy, Acta Crystallogr., Sect. B, 1997, 53, 569.
6. (a) D. Braga, F. Grepioni, E. Tagliavini, J. J. Novoa and F. Mota, New J. Chem., 1998, 22, 755; (b) J. J. Novoa, I. Nobeli, F. Grepioni and D. Braga, New J. Chem., 2000, 24, 5.
7. (a) V. R. Peddireddi and G. R. Desiraju, J. Chem. Soc., Chem. Commun., 1992, 988; (b) D. Braga, D. K. Biradha, F. Grepioni, V. R. Pedireddi and G. R. Desiraju, J. Am. Chem. Soc., 1995, 117, 3156.
8. H.-J. Böhm, S. Brode, U. Hesse and G. Klebe, Chem. Eur. J., 1996, 2, 1509.
9. (a) M. L. Chabinyc and J. Brauman, J. Am. Chem. Soc., 2000, 120, 10863; (b) M. L. Chabinyc and J. Brauman, J. Am. Chem. Soc., 2000, 122, 8739.
10. D. Braga, L. Maini, M. Polito, M. Rossini and F. Grepioni, Chem. Eur. J., 2000, 6, 4227.
11. D. Braga, L. Maini, M. Polito and F. Grepioni, Organometallics, 1999, 18, 2577.
12. (a) G. M. Sheldrick, SHELXL97, Program for Crystal Structure Determination, University of Göttingen, Göttingen, Germany, 1997.; (b) E. Keller, SCHAKAL99 Graphical Representation of Molecular Models, University of Freiburg, Germany, 1999.; (c) A. L. Spek, Acta Crystallogr., Sect. A, 1990, 46, 31.
13. (a) G. R. Desiraju and T. Steiner, The Weak Hydrogen Bond in Structural Chemistry and Biology, Oxford University Press, Oxford, 1999.; (b) D. Braga, F. Grepioni and G. R. Desiraju, Chem. Rev., 1998, 98, 1375.
14. D. Braga, L. Maini, M. Polito and F. Grepioni, Chem. Commun., 1999, 1949.
15. G. A. Jeffrey, An Introduction to Hydrogen Bonding, Oxford University Press, Oxford, 1997, pp. 27–30..
16. (a) D. Braga, F. Grepioni and J. J. Novoa, Chem. Commun., 1998, 1959; (b) D. Braga, L. Maini, F. Grepioni, F. Mota, C. Rovira and J. J. Novoa, Chem. Eur. J., 2000, 6, 4536; (c) D. Braga, F. Grepioni and J. J. Novoa, New J. Chem., 2001, 25, 226.
17. (a) T. Steiner, Chem. Commun., 1999, 2299; (b) M. Mascal, C. E. Marjo and A. J. Blake, Chem. Commun., 2000, 1591; (c) P. Macchi, B. B. Iversen, A. Sironi, B. C. Chakoumakos and F. K. Larsen, Angew. Chem., Int. Ed., 2000, 39, 2719.
18. (a) J. D. Dunitz and A. Gavezzotti, Acc. Chem. Res., 1999, 32, 677; (b) D. Braga and F. Grepioni, Acc. Chem. Res., 2000, 33, 601.
19. (a) L. Leiserowitz and G. M. J. Schmidt, J. Chem. Soc. A, 1969, 2372; (b) L. Leiserowitz, Acta Crystallogr., Sect. B, 1976, 32, 775; (c) Z. Berkovitch-Yellin and L. Leiserowitz, J. Am. Chem. Soc., 1982, 104, 4064; (d) O. N. Ventura, J. B. Rama, L. Turi and J. J. Dannenberg, J. Phys. Chem., 1995, 99, 131; (e) A. Gavezzotti, Chem. Eur. J., 1999, 5, 567; (f) A. Gavezzotti, G. Filippini, J. Kroon, B. P. van Eijck and P. Klewinghaus, Chem. Eur. J., 1997, 3, 893; (g) F. H. Allen, W. D. S. Motherwell, P. R. Raithby, G. P. Shields and R. Taylor, New J. Chem., 1999, 23, 25.
20. F. H. Allen and O. Kennard, Chemical Design Automation News, 1993, 8, 31.
21. R. H. Blessing, Acta Crystallogr., Sect. B, 1986, 42, 613.
22. See for a general entry see Crystal Engineering: from Molecules and Crystals to Materials, ed. D. Braga, F. Grepioni and A. G. Orpen, Kluwer Academic Publishers, Dordrecht, 1999..
23. (a) C. B. Aakeröy, P. B. Hitchcock, B. D. Moyle and K. R. Seddon, J. Chem. Soc., Chem. Commun., 1989, 1856; (b) N. J. Long, Angew. Chem., Int. Ed. Engl., 1995, 34, 21.

---