The Incremental Description of Host-Guest Complexes: Free Energy


The Incremental Description of Host-Guest Complexes: Free Energy...

0 downloads 60 Views 1MB Size

3648

J. Org. Chem. 1993,58, 3648-3653

The Incremental Description of Host-Guest Complexes: Free Energy Increments Derived from Hydrogen Bonds Applied to Crown Ethers and Cryptandsl Hans-Jorg Schneider,'*t Volker Rudiger,+and Oleg A. Raevsky*J FR Organische Chemie der Uniuersitat des Saarlandes, D W 6600 Saarbriicken, Germany, and Institute of Physiologically Active Compounds, Russian Academy of Science, 142432 ChernogolovkalMoscow, Russia Received December 29, 1992

The analysis of literature data for crown ether and cryptand-cation complexes with the principle of additive pairwise interactions yield excellent linear correlations between observed and calculated complex stabilities as long as geometric fit between ligand site and cation is present and no substantial strain is built up during complexation. These conditions are controlled partially by molecular mechanics simulations and partially by comparison to X-ray data. The cornerstone values used for each single ligand-cation interaction are taken from the electron donor parameters ED/Ci which have been derived earlier from over 900 hydrogen bond associations in carbon tetrachloride. The observed electron acceptor abilities EA of the different cations (alkali metal and ammonium ions) are a linear function of, e.g., hydration energies of these ions. Thus, a simple procedure is provided to calculate the stability of over 120 complexes on the basis of one EA parameter for each cation and only 11 ligand ED parameters (same for crowns and cryptands) which moreover are derived from independent sources and do not need to be adjusted. Intermolecular associations are based on a bewildering number of noncovalent binding mechanisms which complicates quantitative structure-activity correlations. Although macrocyclic ionophores represent the oldest and most developed classof host-guest complexes, also in terms of practical applications,z a quantitative understanding of such complexes, or the ability to predict complexation energiesfor given structures as a prequesite for the rational design of ligands, is still in the state of infancy. Although force field, molecular dynamics, as well as free energy pertubation calculations for crown ethers and related hosts were greatly advanced recently? reliable computations of corresponding complexation energies are essentially limited to associations with electroneutral guests: Thus,

Chart I

Table I. ED*and Ik+* Increments IKtb

ED"

CHsOH

2.1

-6.83

Ha0 -5.10

2 3 4 5 6

1.7 1.0 1.1 0.9 0.4

-5.53 -3.25 -3.58 -2.93 -1.30

-4.13 -2.43 -2.67 -2.19 -0.97

7 8 9

0.0 3.5 2.6

-11.38 -8.45

10

0.0

11

3.0

no. 1

fragment

t Saarbroken.

Chernogolovka/Moecow. (1) SaarbrOcken Series on Supramolecular Chemistry. 36. For part 35, see ref 15. (2) (a) Cation Binding by Macrocycles/Complexation of Cationic Speciesby Crown Ethers; Inoue, Y., Gokel, G. W., Eds.; Marcell Dekker, Inc.: New York, Basel, 1990. (b) Crown Ethers and Analogs; Patai, S., Rappoport, Z., Eds.; Wiley: Chichester, 1989. (c) Izatt, R. M.; Pawlak, K.; Bradshaw, I. S.; Bruening, R. L. Chem. Reu. 1991,91,1721-2085. (d) Solov'ev, V. P.; Vnuk, E. A.; Strahova, N. N.; Raevsky, 0. A. Thermodynamics of Complexation of Metals with Cyclic Polyethers (RUBS.); Vinity: Moscow, 1991;pp 1-379. (e) Hancock, R. D.; Martell, A. E. Chem. Rev. 1989,89, 1875. (3) (a) Wipff, G.; Kollman, P. J. Am. Chem. SOC. 1982,104,3249. (b) Wipff, G.; Kollman, P. Nouu. J. Chem. 1985, 9, 457. (c) Wipff, G. J. Coord. Chem., in press. Boudon, S.; Wipff, G. J . Chim. Phys. 1991,88, 2525. Auffinger, P.; Wipff, G. J. Am. Chem. SOC. 1991,113,5976. Gehin, D.; Kollman, P. A.; Wipff, G. Ibid. 1989,111,3011. (These and earlier papers emphasize also the role of water and the preorganization of ionophore ligand as well as asymmetries in energy minimized structures. We thank Prof. Wipff for pre- and reprints.) (d) Grootenhuis, P. D. J.; Kollman, P. A. J. Am. Chem. SOC. 1989,111,2152 and earlier papers. (e) Strantema, T. P.;McCammon, J. A. J. Chem.Phys. 1989,91,3631. Mazor, M. H.; McCammon, J. A.; Lybrand, T. P. J. Am. Chem. SOC. 1990,112, 4411 (one example of relatively satisfadory results with FEP methods). (0Van Eerden, J.; Harkema, S.;Feil, D. J. Phys. Chem. 1988,92,5076. (g) Lifson, S.; Felder, C. E.; Shanzer, A.; Lilman, J. In Progress in Macrocyclic Chemistry; Izatt,R. M., Christensen, J. J., Eds.; Wiley: New York, 1987;Vol. 3, pp241-308. (h)Bradshaw,J. S.;Huszthy, P.; McDaniel, C. W.;nu,C. Y.;Dalley, N. K.; Izatt, R. M.; Lifson, S. J. Org. Chem. 1990,50,3129. (i) Fukushima, K. Bull. Chem. SOC. Jpn. 1990,63,2104. (j) Dang, L. X.;Kollman, P. A. J. Am. Chem. SOC. 1990,112,5716. (k) Zomppa, Inorg. Chem. 1986, 25, 1527 and references cited in ref 2a-e. t

0.00

0.00

-9.75

0.00 -8.51 -6.32 0.00 -7.29

Electron Donor parameters, derived from hydrogen bond associations in CC4; see text; the asterisk denotes the binding element. Incrementa for potassium complexation (kJ/mol) (for other cations, see Table 111). R = alkyl.

Kollman, Wippf, and others have, e.g., demonstrated early the dependence of calculated complex stabilities on the (4) (a) Damewood, J. R.; Anderson, W. P.; Urban, J. J. J. Comput. Chem. 1988, 9, 111. (b) Grootenhuis, P. D. J.; Kollman, P. A. J. Am. Chem. SOC. 1989,111, 4046. (c) Grootenhuis, P. D. J.; van Eerden, J.; Dijkstra, P. J.; Harkema, S.; Reinhoudt, D. N. Zbid. 1987, 109, 8044.

QQ22-3263/93/1958-364~~Q4.QQlQ 0 1993 American Chemical Society

J. Org. Chem., Vol. 58,No. 14, 1993 3649

Description of Host-Guest Complexes Scheme I Rl

Ra

I

I1

XI

"0

L o 9

Xlll

XIV

IV

g$

OR

xv

XVI

XVll

XVlll

VI

VI1

Vlll

IX

X

IX

xx

chosen permanent and induced charges.& We calculate, e.g., up to 10%changes of the potassium-oxygen distances by rather arbitrary choices of the used dielectric constant between 1 and 3 D (see below). It would be of considerable advantage if the noncovalent forcesin ionophore complexes could be predicted by simple additive schemes reminiscent of the long-standingLFERtype description for covalent bond making and breaking. Such attempts were made many years ago by Cram et al.? but with eight parameters for the description of nine

complexes the results of this early attempt were not very encouraging. The Saarbriicken group hae recently shown that the free complexation energy AGt in many supramolecular complexes can be partitioned in singlecontributions by linear correlations between AGt and the s u m of single interactions.6 In a typical ionophore complex we may encounter, e.g., three different interactions of a metal cation M+and binding sites A, B, and C (representing, e.g., alkyl-0, aryl-0, and N). For such a complex (Chart

(5) (a) Timko, J. M.; Moore, S. S.;Walba, D. M.; Hiberty, P. S.;Cram, D. J. J. Am. Chem. SOC.1977,99,4207. For related approaches see also ref 2e and (b) files, T. M. In ref 2a, pp 203 ff. (c) Behr, J. P.; Lehn, J.-M.;Vierling, P. Helu. 1982, 65, 1853.

Znt. Ed. Engl. 1991,30,1417. (b) Schneider, H.-J.;Theis, I. Zbid. 1989, 101, 757; Angew. Chem., Znt. Ed. Engl. 1989, 28, 753. (c) Schneider, H.-J.;Blatter, T. Angew. Chem. 1992,104. (d)Schneider,H.J.;Schieatel,

(6) (a) Schneider, H.-J.Angew. Chem. 1991,103,1419; Angew. Chem.,

T.; Zimmermann, P. J. Am. Chem. SOC.1992,114,7698.

3650 J. Org. Chem., Vol. 58, No. 14, 1993

I) the total AGt would be the sum of the individual increments: AG, = nAAGA nBAGB + ncAGc + ... In our earlier analyses we have obtained the individual increments AGA etc. directlyfrom the correlations,leading, e.g., to a surprisingly constant value of AG = 5 f 1kJ/mol per single salt bridge in now more than 50 different ion pair complexes,6 which range from zinc sulfate6dto DNA/ polyammonium interactions.& In the present analysis of ionophores we choose another approach and set out to use increments for the different binding sites from independent sources. These increments which should quantify the interaction of a single ligand site and a cation obviously must reflect primarily the ligand electron donor capacity. Such numbers-referring, however, to Lewis base complexes with AH instead of AG values-are available from the extensive studies by Gutmann et al.? as well as by Drago et al.,8 the latter approach being based on three terms accountingfor electrostatic,covalent,and CT interactions. In related attempts Abraham, Kamlet, Taft et al. have correlated hydrogen bond complex free energies with acidities and basicities of the c~mponents.~ The Chernogolovka/Moscowgroup of the present authors, starting from earlier proposals by Iogansen et al.,l0has derived AH and AG increments from a large body of thermodynamic data for single hydrogen bond complexes based on eq 2 which allows to extract electron-withdrawingand -donating energyfactors C, and Cb from many 1:1equilibriain carbon tetrachloride.11J2

Schneider et al.

+

AGt = 2.43CaC,

+ 5.70

(2) The electron-donating increments to be used for the ionophores, later called ED factors, are listed in Table I; they have been found to represent on a uniform scale the minimum set of parameters for good correlations (coefficients r 1 0.980 usually) for more than 900 hydrogen bond complexes.11J2 One incentive of this study was to see whether parameters derived, e.g., from such hydrogen bond complexes in carbontetrachloride,11J2are of sufficient fundamental significancefor the prediction of ionophorecation complex stabilities in protic solvents. Complexation energies for a large and representative series of ionophores (Scheme I) with alkali metal cations and tentatively also with ammonium ions were assembled (7) (a) Gutman, V. The Donor-Acceptor Approach to Molecular 1nteraction.s;Plenum Press: New York, 1978. (b) Gutman, V.; Resch, G. Stud. Phys. Theor. Chem. 1983,27 (Ions and Molecules in Solution), 203. (8) (a) Drago, R. S. Structure and Bonding;Springer-Verlag: Heidelberg, 1973; pp 73-139. (b) Drago, R. S.;Vogel, G. C.; Needham, I. E. J. Am. Chem. SOC.1971,93,6014. (c) Kroeger, M. K.; Drago, R. S. J. Am. Chem. SOC.1981,103,3250. (9) (a) Kamlet, H.-J.; Taft, R. W. J. Am. Chem. SOC.1976, 98, 377, 2886. (b)Abraham, M. M.; Duce, P. P.;Schulz,R. A., et al. J. Chem.SOC., Perkin Trans. 1 1986,1501. (c) Abraham, M. M.; Grellier, P. L.; Prior, D. V.; Taft, R. W.; Morris, J. J.; Taylor, P. J.; Laurence, C.; Berthelot, M.; Doherty, R. M.; Kamlet, M. J.; Abbound, J.-L. M.; Sraidi, K.; Guihheuf, G. J . Am. Chem. SOC.1988,110,8534. (10) (a)Iogansen,A. V. Dokl. Acad. Nauk SSSR 1965,164,610(Russ.); Chem. Abstr. 1966, 64, 2868h. (b) Iogansen, A. V. Teor. Eksp. Khim. 1971, 7,302,312 (Russ.); Chem. Abstr. 1971, 75,101848m, 101824a. (11) (a) Raevsky, 0.A.; Novikov, V. P. Khim.-Farm.Zh. 1982,16,583; Chem. Abstr. 1982,97, 33128d. (b) Raevsky, 0. A.; Grigor’ev, V. Yn.; Solov’ev, V. Khim.-Farm. Zh. 1984, 18, 578; Chem. Abstr. 1985, 103, 8380711. (c) Raevsky, 0. A.; Grigor’ev, V. Yn.; Solov’ev, V. Khim.-Farm. Zn. 1989,12,1294 (Russ.);Chem.Abstr. 1990,112,115266g. (d)Raevsky, 0. A. Russ. Rev., Eng. Ed. 1990,59,219-233. (12) (a)Raevsky, 0.A.; Grigor’ev,V. Yn.; Kireev, D. B.; Zefirov, N. S. B u n t . Struct.-Act. Relat. 1992,11, 49. (b) Raevsky, 0. A.; Grigor’ev, V. Yn.; Kireev, D. B.; Zefirov, N. S. J. Chem.Phys. Chem.Biol. 1992,89, 1747.

Figure 1. Plots for the l&crown-G/+NH, complex from energyminimized structures,showing the DMsymmetry. Shaded circles indicate oxygen atoms.

from the literature2 (Table I*, supplementary material). Taking into consideration the dependence of the complexation free energy AGt also on cation, anion, and ligand desolvation/solvation as well as on the ~ o l v e n t . ~only J~ complexes with the same anion (mostly chloride) and the same solvent (mostly methanol) were used for the correlations. The application of pairwise interaction increments rested until now on the presence of sufficient contacts between the individual binding sites in host and guest.6 Therefore, the complex structures used in the correlations were checked to meet this condition (e.g., mismatch of distancesbetween metal and ligand atoms