Journal of Molecular Structure 751 (2005) 172–183 www.elsevier.com/locate/molstruc
peri-Interactions in naphthalenes, 13 [1] 8-Dimethylamino-naphth-1-yl carbinols as model systems for potential N/Si/P interactions Paulina M. Dominiaka, So¨nke Petersenb, Berthold Schiemenzc, Gu¨nter P. Schiemenzb,*, Krzysztof Wozniaka,* a
Chemistry Department, Warsaw University, 02 093 Warszawa, ul. Pasteura 1, Poland b Institute of Inorganic Chemistry, University of Kiel, D-24098 Kiel, Germany c Institute of Inorganic Chemistry, University of Heidelberg, Heidelberg, Germany Received 7 March 2005; accepted 26 April 2005 Available online 14 July 2005
Abstract In (8-dimethylamino-naphth-1-yl) (‘DAN’) carbinols DAN–C(OH)R1R2, the N atom can approach the HO group unto N/H-O hydrogen bond distance only when the steric conditions are favourable. The energy gain of such N/H–O interactions is insufficient to force R1 and R2 into otherwise unfavourable conformations. The geometry of the naphthalene may cause N, O and C atoms to reside in positions similar to those typical for hydrogen bonds though no N/H–O and N/H–C hydrogen bonds may actually be involved. By analogy, it seems unlikely that in peri-donor/acceptor substituted naphthalenes D–C10H6–A dative interactions (D/A) of similar or less energy as such N/H-O interactions can interfere with the geometry conserving forces of naphthalene and the steric effects of the peri substituents. q 2005 Elsevier B.V. All rights reserved. Keywords: peri-Substituted naphthalenes; Intramolecular hydrogen bonds; Weak donor/acceptor interactions; Steric effects
1. Introduction By definition, the sum of the van der Waals radii of two atoms X and Y, Sr(vdW)[X, Y], is the interatomic distance d(X/Y) at which repulsive forces and attractive van der Waals forces are equal [2,3]. Sr(vdW) presently serves as a measure of the minimum interatomic distance in the absence of other attractive forces. There is a host of different attractive forces which are stronger than the notoriously weak van der Waals forces [3–6] and thus cause d(X/Y)!Sr(vdW) [3,5]. It is, therefore, not possible to conclude from experimentally found d(X/ Y)!Sr(vdW)[X, Y] that a particular attractive force, such as a dative [two electron] bonding interaction D/A between a s-electron donor D and an electron acceptor A, is operative [3,6]. There exist even non-attractive forces which easily put X and Y (including D and A) into sub-van der Waals distances [2,3]. One of them is the geometry* Corresponding authors. Tel./fax: C48 22 8222892/C49 431 8802441. E-mail addresses:
[email protected] (G.P. Schiemenz),
[email protected] (K. Wozniak).
0022-2860/$ - see front matter q 2005 Elsevier B.V. All rights reserved. doi:10.1016/j.molstruc.2005.04.042
preserving force of the naphthalene skeleton which tends to be planar and to have bond angles of 1208 throughout [2,3, 5–12]. These geometrical features would place equal substituents X in the peri positions at a distance of d(X/ X) Z246.8 pm and unequal substituents X and Y at a distance of d(X/Y)Z ca. 247–255 pm, e.g. d(N/Si) Z 251 pm (’peri distance‘, PD) [3,5–11,13]. This is much less than Sr(vdW), e.g. Sr(vdW)[N, Si]Z345 pm [14]. Steric repulsion between X and Y generally forces the naphthalene skeleton into distortion and leads to an increase of d(X/Y) which, however, consistently remains much below Sr(vdW)[X, Y] (e.g. d(N/Si) typically ca. 265–300 pm [15,16]). On the other hand, the naphthalene geometry yields quite easily to strong attractive forces such as the bond length preserving forces exerted by covalent bonds X– Y, e.g. H2C–CH2 in acenaphthene, d(C–C) Z154 pm [17], including ’dative‘ and even hypercoordinate bonds between D and A, such as N/P, d(N–P) Z213 pm [10]. Hence, both to steric repulsive and covalent bond attractive forces, the naphthalene skeleton is the loser. For a dimethylamino group as a (mediocre [12]) s-electron donor and a silicon or phosphorus atom as a potential - though doubtful - electron acceptor in 8-dimethylamino-naphth-1-yl (’DAN‘) silanes and phosphines, the concept of ’a weak dative N/Si/P
P.M. Dominiak et al. / Journal of Molecular Structure 751 (2005) 172–183
bonding interaction‘ has been promulgated on the basis of d(N/Si/P)!Sr(vdW)[N, Si/P] though [Sr(cov)[N, Si/ P] (r(cov)Zcovalence radius) [16,18–20]. A peculiar property of these alleged weak bonding interactions is that unlike conventional covalent bonds they are stretchable ad libitum and can easily assume any bond length between Sr(cov) and Sr(vdW). A linear bond order equation between 1 (Sr(cov)) and 0 (Sr(vdW)) has been devised [21] and applied to DAN-A systems [22]. In our opinion, this very concept seems questionable [2,3,5,10]. At best, only a small bond energy should be associated with a weak bonding interaction [3,6], and the question arises how the naphthalene system would respond to it: would it still yield to the ’bond‘, or would it rather dictate the conditions of such interatomic interaction? Some insight might be expected from a study of hydrogen bonds. Their bond energies are much smaller than those of covalent bonds. Like the latter, they prefer certain interatomic distances, but they are more amenable to bond length alterations [11], and their bond lengths are within the same range of interatomic distances as those of the alleged N/Si/P bonds. Their potential impact on the naphthalene geometry may therefore be expected to be comparable. The most prominent DAN compounds to form hydrogen bonds are the protonated proton sponges, first of all 1,8bis(dimethylamino)naphthalene (1a, see Scheme 1) after protonation. For our purpose, however, these compounds are unsuitable, because d(N/N) of strong N/H–NC hydrogen bonds, occasionally as short as 253 pm [23], 256–261 pm in 1a$HC[24]), 255 pm in 1b$HC[25], is not significantly different from PD [26] and therefore neither able to deform the C10 skeleton nor prone to be influenced by the geometry of the latter. We therefore turned to carbinols DAN-C(OH)R1R2 (2; see Scheme 1, Fig. 1 and Table 1). In their ideal geometries, the bond lengths and bond angles (1208 at sp2–C, 109.58 at Me2 N
RO R3 8
5
1
R
1
X
D
2
R
+ R1 P R2
2 3
6
R
–
R1 C
9
7
R
sp3–C) would permit the O atoms to approach the N atoms unto distances typical for N/H–O hydrogen bonds of moderate strength (or even less). Rotation around the periC–C bond would give rise to various N/O distances and, in case of N/H–O bond formation, to various bond energies of the hydrogen bonds. These energies might have an impact both on the geometry of the C10 skeleton and on the conformations of the peri-substituents, mainly concerning rotation of the –C(OH)R1R2 group around the peri-C–C bond. In view of Steiner’s caveat that not all short O/H–C contacts are hydrogen bonds [27], attention must focus on the possibility that the molecular geometry imposes to the N, O and H atoms interatomic distances and N/H–O angles which are compatible with N/H–O hydrogen bonds but which, unlike the situation in the protonated proton sponges, are not associated with their typical bond energies and hence do not qualify for hydrogen bonds in the proper sense. This could be the case when the N-lone pair is not directed towards the H-O bond or, in non-linear N/H–O alignments, does not lie in the N/H–O triangular plane. In DAN–OH, the fact that it does is indicated by its structure with a nearly planar C10 skeleton, N, O and H residing nearly in the C10 plane and the N-methyl groups nearly symmetrically above and below it in anticlinal (ac) positions [28] with respect to the C(1)/C(8) connecting line [29]. Deviations of the Me2N group from this symmetrical conformation would be indicative of noncoplanarity of the lone pair with the C10 plane. When this project started there was only one crystal structure of a peri-carbinol deposited in CSD [7]. In the meantime one of the structures we use in our study, 2a/b (see Scheme 1 and Fig. 2), has already been published [30]. However, as it was not discussed in the context of steric interactions, we decided to incorporate our results for this compound into this work.
4
NMe2
173
10
4
2, 3 R1
R2
R3
R4
4
D
X
R1
R2
H (R)
4a Me2N
O
any
any
DAN Ph
1a H
2a
H
Ph
Me2N
1b OMe
2b
Ph
H
Me2N
H (S)
4b Me2N
S
1c NMe2
2c
DAN
H
Me2N
H
4c Ph2P
Se Ph
2d
c-C3H5 c-C3H5
Me2N
H
2e
CHMe2 CHMe2
Me2N
H
3
CMe3
H
Me
CMe3
Scheme 1. Definition of compounds.
Ph
174
P.M. Dominiak et al. / Journal of Molecular Structure 751 (2005) 172–183
Fig. 1. ORTEP plots showing the labelling of atoms and atomic displacement parameters. On the right side of this figure, the projections of the compounds along the C(9)–C(10) bond are shown.
2. Results and discussion 2.1. Bis(8-dimethylamino-naphth-1-yl)carbinol (2c) In principle, 2c belongs to the compound type A2XBC with a tetrahedral center X in which the A substituents are enantiotopic. However, hindered rotation around the peri bonds as met frequently in molecules (DAN)XR1R2R3 even in solution [7,8,31], renders the DAN groups of (DAN)2XR1R2 diastereotopic, i.e. different. In 2c, the DAN groups are labeled DANI and DANII. The atoms and groups within DANI,II bear the same label, and the carbinol carbon atom is called Ccarb. The geminal methyl groups of each
Me2N group are labelled MeA and MeB, and the sides of the respective naphthalene system at which they are situated, as faces A and B. For the conformations of MeA,B and the substituents at Ccarb with respect to the C(1)/C(8) connecting line of the C10 system, we use the terminology of Klyne and Prelog [28]. In a previous study of solid 2c, the vicinity of the N atom and the HO group had induced us to assume a N/H-O hydrogen bond [7]. However, closer inspection now revealed that the molecular geometry provides the required interatomic distances only in DANI, and even there imposes significant restrictions. In DANI, the NI/Ccarb distance, 287.6 pm (see Table 2), is shorter than Sr(vdW)[N, C]Z325 pm [14] by 37.4 pm.
P.M. Dominiak et al. / Journal of Molecular Structure 751 (2005) 172–183
175
Table 1 Crystal data and structure refinement details for 2a/b and 2d Identification code
DAN-C(OH)(Ph)H (2a/b)
Empirical formula Formula weight (g/mol) Color Crystal size (mm3) Crystal system Space group ˚ , deg) Unit cell dimensions (A
C38H38 N2 O2 554.7 Transparent 0.15!0.20!0.20 Triclinic P-1 aZ8.521(2), aZ76.52(3) bZ10.551(2), bZ81.77(3) cZ17.737(4), gZ89.95(3) 1533.9(6) 2 592 1.201 0.71073 0.074 K11%h%11K14%k%12 K22%l% 3
˚ 3) Volume (A Z F(000) Density (calculated) (mg/m3) ˚) Wavelength (A Absorption coefficient (mmK1) Index ranges
Theta range for data collection (deg) Diffractometer Reflections collected Independent reflections Temperature (K) Final R indices (IO2s(I)) R indices (all data) Absolute structure parameter Extinction coefficient Weight (PZ(Max(Fo2,0)C2*Fc2K/3)
3.6–28.8 KUMA 4 CCD 11079 6930 (RintZ0.0266) 293 R1Z0.051, wR2Z0.132 R1Z0.100, wR2Z0.152 0 0.00 1/[s2(Fo2)C(0.0879*P)2]
˚ K3) Largest diffraction peak and hole (e A
0.17 and K0.15
With a typical covalent single bond length, d(N–C(sp3))Z 147 pm [6], this value formally corresponds to 21% hypercoordination character of Ccarb according to the procedure of Holmes [21]. This calculation serves the purpose of demonstrating the inapplicability of this method in this case and, consequently, in all analogous cases [2,3,5, 10]. On the other hand, d(NI/Ccarb) exceeds PD by ca. 41 pm; this is a measure of steric repulsion. The latter shows up also in the splay angle, i.e. the three bay angles K3608 [2,3,5,10],C10.18. As in many other cases, the major burden of distortion is carried by the inner angle, C(1)I– C(9)I–C(8)I, 127.58, whereas the angle NI–C(8)I–C(9)I, 119.88 is not affected. This widespread phenomenon invalidates rationalizations which focus on N–C(8)–C(9) and leave C(1)–C(9)–C(8) out of consideration [32]. As a consequence of C(1)I–C(9)I–C(8)I exceeding 1208, the distance d(C(1)I/C(8)I)Z259.7 pm is larger than PD by DdZ ca. 13 pm. This enlargement is counterbalanced by a slight forshortening of d(C(4)I/C(5)I) at the other side of the CI10 skeleton (dZ243.3 pm, DdZ ca. K4 pm). In DANII, all data are quite similar (d(NII/Ccarb)Z 286.6 pm, C(1) II –C(9) II–C(8) IIZ126.28, N II –C(8) II– C(9)IIZ120.18, splay angle C10.88, d(C(1)II/C(8)II)Z 256.6 pm, d(C(4)II/C(5)II)Z243.7 pm). The influence of
DAN-C(OH)(C3H5)2 (2d) C19H23 N1 O1 281.4 Transparent 0.25!0.15!0.10 Orthorhombic P212121 AZ8.7940(18) bZ10. 150(2) cZ17.064(3) 1523.1(5) 4 608 1.227 0.71073 0.075 K11%h% 8 K13%k%13 K22%l%2 3.3–28.5 KUMA 4 CCD 10599 3610 (RintZ0.064) 293 R1Z0.053, wR2Z0.135 R1Z0.102, wR2Z0.149 0.0(2) 0.013(3) 1/[s2(Fo2)C(0. 0902*P)2] 0.19 and –0.18
Fig. 2. Intra- and intermolecular H-bonds in DAN–C(OH)(Ph)H (2a/b).
176
P.M. Dominiak et al. / Journal of Molecular Structure 751 (2005) 172–183
Table 2 Selected most important structural parameters of 2a–d (distances in pm, angles in degrees) Structural parameter
2c: DANI
2c: DANII
2b
2a
2d
N1/Ccarb* N1-C8 Ccarb-C1 C1/C8 C4/C5 Ccarb–O1 N1–C8–C9 Ccarb–C1–C9 C1–C9–C8 N1–C8/C1–Ccarb C81–N1–C8/C1 C82–N1–C8/C1 C21–Ccarb–C1/C8 C31–Ccarb–C1/C8 O1-Ccarb–C1/C8 2e(N1)**–N1/Ccarb
287.6 144.5 153.5 259.7 243.3 144.3 119.76 122.85 127.47 8.57 86.96 K148.45 K58.43 (H) K178.34 60.96
286.6 144.5 152.6 256.6 243.7 143.6(2) 120.07 124.49 126.24 9.17 108.28 K126.63 73.82 K47.44 (H) K165.22
287.6(2) 144.1(2) 153.1(2) 258.3(3) 244.0(4) 143.9(2) 120.0(2) 123.6(1) 127.1(1) 7.3(1) 89.9(2) K143.2(2) K62.0(1.0) (H) K179.8(2) 56.7(1) 43.4
284.8(3) 142.0(2) 152.2(2) 255.0(3) 244.2(4) 141.7(3) 119.1(2) 123.0(1) 125.4(2) 23.9(1) 80.0(2) K149.7(2) 74.7(2) K52.7(1.0) (H) K164.6(1) 47.7
314.1(3) 145.3(3) 155.4(3) 262.3(3) 239.6(4) 122.8(2) 127.2(2) 129.5(2) 6.5(2) K119.4(2) 116.5(2) 73.9(2) K165.0(2) K46.7(3) 46.5
*CcarbZC11 in the case of 2a/b and 2d; ** 2e(N1)Z electron pair at N1.
the NI/H–O interaction on these parameters, then, is insignificant. In DANI, MeAI is in a position virtually perpendicular to the CI10 plane (dihedral angle MeAI–NI–C(8)I/C(1)I 87.08), MeBI in a borderline position of the anticlinal (ac) and the antiperiplanar (ap) sectors at the opposite face (face B) of CI10 (dihedral angle MeBI–NI–C(8)I/C(1)I K148.58). In DANII, the conformation of Me2NII is similar, though less dissymmetric with regard to the CII10 plane (dihedral angles MeA,BII–NII–C(8)II/C(1)II C108.3 and K126.68, i.e. both ac). This type of conformation is a general feature of DAN compounds except for cases with a bond between N and the atom bound to C(1) [10]. It implies that in both DANI and DANII the lone pair at N is pointing into the syn region though to a position outside the C10 plane. The respective conformations of the Me groups suggest that the angle between the lone pair and the C10 plane is 30.88 (borderline between synperiplanar (sp) and synclinal (sc)) in DANI, but much smaller in DANII (9.28, sp). In DANI, the lone pair and the HO group reside on the same face of the CI10 plane. The almost universal occurrence of similar conformations of the Me2N group permits to conclude that in DANI, this arrangement is not a consequence of a hydrogen bond. The conclusion is in accordance with the fact that the lone pair (in its assumed direction) does not point towards the hydrogen atom. With respect to both DAN groups, one of the three bonds between Ccarb and the substituents H, HO and the second DAN is nearly in the C10 plane and directed ’outwards‘ (i.e. in the ap region, nearly ecliptic with the C(1)–C(2) bond [33]); the other two are above and below the C10 plane and in the syn region, thus creating an imperfect herring-bone pattern for the Me–N bonds and two of the bonds emanating from Ccarb. Due to the unsymmetrical position of MeA and MeB, the two faces of the naphthalene skeleton differ in geometry and consequently in the size of the peri space available for a substituent at Ccarb. As a matter of fact, for
both DANI and DANII, face A is occupied by the H atom, i.e. the smaller peri space by the smallest substituent at Ccarb. In DANI the ap position is occupied by DANII and the syn position of face B by OH (dihedral angle O-CcarbC(1)I/C(8)I 61.08, sc). The C10 planes of DANI and DANII are virtually orthogonal to each other - a conformation which is obviously sterically much more favourable than a coplanar arrangement. Thus, the size of the substituents seems to be the decisive factor which determines the choice between the various conformations to which rotation around the C(1)-Ccarb bond might give rise. In the X-ray structure determination, the H atom of the HO group has been located. d(N/O) Z273.8 pm, d(N/H) Z188.2 pm, double as long as d(O-H) Z93.5 pm, and the angle N/H-O, 1518, are formally compatible with a hydrogen bond much weaker than a nearly linear and much shorter N/H-O hydrogen bond which Steiner et al. recently have classified as ’a very short O-H/N hydrogen bond‘ (d(N/O) Z258.8 pm, d(N/H) Z153.5 pm, only 44% longer than d(O-H)Z106.8 pm; see Table 3) [34]. According to these criteria, the N/H-O interaction in 2c may represent mainly electrostatic interactions, with only a small charge transfer contribution [35], so that the unfitting direction of the lone pair may not play a prohibitive role. On the other hand, the bond energy of this N/H-O interaction would not be expected to play a significant role either. With respect to DANII, the positions of the HO group and the other DAN group are interconverted (dihedral angles C(1)I-Ccarb-C(1)II/C(8)II 73.88, sc; O-Ccarb-C(1)II/C(8)II 165.28, ap). Due to its ap position, the HO group is unable to engage in a NII/H-O hydrogen bond. The different behaviour of the two DAN groups towards N/H-O interaction thus is recognized to be rooted in stereochemistry. In the conformation in which the smallest substituent, viz. H, resides in the smallest pocket, NI is in close contact with the H-O group, and NII is not. However, the geometry does not permit the N-lone pair to approach
P.M. Dominiak et al. / Journal of Molecular Structure 751 (2005) 172–183
177
Table 3 The closest short contacts and hydrogen bond parameters (distances in pm, angles in degrees) Compound I
2c: DAN 2c: DANI 2c: DANII 2b 2b 2a 2a-DANII/2b–DANI 2d
Donor–H.Acceptor I
O –H/N Ccarb–H/N Ccarb–H/N OI–H/N Ccarb–H/N Ccarb–H/N OII–H/OI O–H/N
Donor–H
H–Acceptor
Donor–Acceptor
H-bond angle
93.5 103.7 103.7 109(3) 107(2) 101(2) 105(3) 103(3)
188.2 252.6 233.6 164(3) 260(2) 238(1) 174(3) 162(3)
273.8 287.6 286.6 264.9(2) 287.5(3) 284.8(2) 278.4(2) 261.2(3)
151.0 99.0 110.4 152(2) 94(1) 108(1) 171(2) 162(2)
the HO group in a favourable direction, so that the N/H-O interaction does not contribute sufficiently to the energy balance of the system to interfere with the steric factors which determine the conformation. Bound to a carbon atom bearing two aryl groups and an electronegative substituent (viz HO), the H atom at Ccarb is somewhat acidic. In principle, therefore, it fulfils the prerequisite for a N/H-C hydrogen bond [36,37]. This hydrogen atom has been located, too. The distances d(N I/H) Z252.6 and d(N II /H) Z233.6 pm are significantly shorter than typical N/H distances in intermolecular N/H-C(sp3) hydrogen bonds (260– 298 pm [37]), and the distances d(N I/C carb) Z 287.6 pm, d(NII/Ccarb) Z286.6 pm fall short of d(N/ C(sp3))Z352–403 pm [37] even more drastically. In such N/H-C(sp3) hydrogen bonds, N/H-C angles ranging from 133–1768 have been observed [37]. In 2c, the corresponding angles NI,II/H-Ccarb are much smaller (99.08, 110.48). Finally, the H atom and the lone pair of NI (in its assumed direction) do not point towards each other. In that respect, NII is somewhat better, but the unfitting distances and angles indicate that the proximity of N and H is caused by the rigid geometry of the molecule rather than by a hydrogen bond. The result is in agreement with Steiner’s aforementioned caveat [27]. 2.2. (8-Dimethylamino-naphth-1-yl)(phenyl)carbinol (2a/b) In order to test our rationalization further, we turned to the carbinol (R,S)-DAN-C(OH)(Ph)H (2a/b). Substitution of DANI by Ph in 2c leads to (R)-DANII-C(OH)(Ph)H (2a), substitution of DANII to (S)-DANI-C(OH)(Ph)H (2b). The formally enantiotopic DAN groups no longer reside in the same molecule, but one in each enantiomer. The X-ray structure determination revealed that the compound crystallizes as a ‘racemate’; there is one molecule of each ‘enantiomer’ in the asymmetric part of the unit cell. 2a and 2b adopt quite different conformations so that they are, in fact, diastereomers. Except for the conformations, both structures are very similar, as well to each other as to that of 2c (see Table 2 for comparison). Correspondingly, the N atom and the HO group are at close contact only in 2b. With respect to 2c, d(NI/O) Z 264.9 pm is shorter by DdZ9 pm, and the H atom is
positioned less unsymmetrically between O and N. Formally, these data bring the N/H-O contact closer to what Steiner et al. [34] had called ’a very short O-H/N hydrogen bond‘ (vide supra), but the angle N/H-O is not increased. Like HO in 2c with respect to Me2NII, the HO group in 2a is not in a suitable position to engage in an intramolecular hydrogen bond. On the other hand, its sterically ’open‘ ap position enables it to form an intermolecular O/H–O hydrogen bond with the oxygen atom of a neighbouring molecule of 2b (see Table 3; Fig. 3). Its interatomic distances resemble those of the N/H–O contacts, but it is closer to linearity (angle O/ H–O 1718). Comparison of the data of 2a–c (cf. Table 2) reveals that again the N/H–O interaction is unable to interfere with the overall structure of the naphthalene system. In 2b and DANI of 2c, both peri substituents adopt almost identical conformations. For 2a and DANII of 2c, this is also true for the carbinol substituent, while the Me2N group exhibits an unusual conformation: With respect to the C10 plane, MeA is no longer in the orthogonal position, but clearly in the sc sector, MeB at the borderline of the ac and ap sectors, and the N-lone pair (in its assumed direction) in the sc sector. Increased steric hindrance between the (Ccarb)H atom and MeA is avoided by a much larger deviation of N and Ccarb from the C10 plane at opposite faces (dihedral angle Ccarb– C(1)/C(8)–N 23.98; ca. 7–88 in 2b, c). While the reason for the different behaviour remains unknown, it is clear that no N/H–O interaction is involved. The parameters of the N/H–Ccarb relation in 2b and 2a are similar to those of DANI and DANII in 2c, respectively (see Table 3). Since the proximity of N and H is a geometrically enforced one, there are again no experimental criteria in favour of a N/H–C hydrogen bond. On the contrary, d(N/H) Z260/238 pm, d(N/ Ccarb)Z287.5/284.8 pm (both too short for a N/H–C bond) and the small angle N/H–Ccarb, 94/1088 (2b/2a), are not in accordance with such a hydrogen bond. It is interesting to note that the involved atoms can go to subhydrogen bond length distances with no obvious difficulty when such positions are favoured by the geometry of the system.
178
P.M. Dominiak et al. / Journal of Molecular Structure 751 (2005) 172–183
Fig. 3. Projection of the crystal lattice of DAN–C(OH)(Ph)H (2a/b) and DAN–C(OH)(C3H5)2 (2d).
2.3. Di(cyclo-propyl)(8-dimethylamino-naphth-1-yl)carbinol (2d) If the lock and key relation of a small steric pocket and a small substituent at Ccarb (viz H) is indeed the decisive factor to determine the conformation, then it was worthwhile to study a carbinol in which H has been replaced by a group larger than HO. For the sake of simplicity, the two groups attached to Ccarb in addition to DAN and HO, preferentially were identical. iso-Propyl groups would be sufficiently bulky, but the corresponding carbinol 2e could be obtained only in low yield: We prepared the carbinols 2 by the reaction of DAN-Li with OZCR1R2. However, easily enolizable ketones such as di(isopropyl)ketone preferentially protonated DAN-Li. On the other hand, di(cyclo-propyl)ketone reacted smoothly as an electrophilic carbonyl compound so that 2d was easily at hand. In it, the cyclo-propyl groups are formally enantiotopic and in case of restricted rotations, diastereotopic. The gross structure of 2d displays all criteria of strong steric hindrance (cf. Table 2). d(N/Ccarb) Z314.1 pm is only ca. 11 pm shorter than Sr(vdW)[N, C] [14]. All three bay angles exceed 1208, N–C(8)–C(9) significantly, CcarbC(1)–C(9) and C(1)–C(9)–C(8) strongly. The splay angle, C19.58, is almost double as large as in 2b and in 2c and 2.6
times larger than in 2a. Nevertheless, the C10 skeleton is almost planar, and the peri substituents are almost in the C10 plane (dihedral angle Ccarb–C(1)/C(8)–N 6.58). In 2d, the smallest group at Ccarb is HO. Not surprisingly, then, HO occupies one of the sc positions, leaving the other sc position and the ap position to the C3H5 groups. However, the Ccarb substituent has its closest counterpart not in 2b and DANI of 2c, but in 2a where the positions of HO and the hydrocarbon substituent are interconverted. All three dihedral angles are almost identical with the respective angles in 2a. Replacement of H in 2a–c by a bulkier substituent (viz HO) might be expected to force MeA from the nearly perpendicular into a skew position. The result would be a less dissymmetric arrangement of the Me2N group with respect to the C10 plane. In 2d, this is indeed the case. The dihedral angles MeA,B-N–C(8)/C(1) are almost equal: 116.5 and K119.48. This should imply that the lone pair at N is in the sp sector almost in the C10 plane, formally deviating from it by 1.58 at face B, while the HO group resides in the sc sector of face A, hence on the opposite side. Since Ccarb is virtually in that plane, too, and d(Ccarb– C(1))Z155.4 pm is ca. 10 pm longer than d(N–C(8))Z 145.3 pm, this means that the lone pair points nicely towards Ccarb though there is no reason to assume a N/Ccarb bonding interaction. This feature invalidates one of the arguments promulgated in favour of a donor–acceptor
P.M. Dominiak et al. / Journal of Molecular Structure 751 (2005) 172–183
interaction in DAN compounds such as DAN-phosphines (N/P) [20,38]. It emerges that the steric interaction of substituents determines the direction of the lone pair, and not vice versa. This conclusion is supported by the structures of 1a–c [25, 39] and a series of 1,8-bis(diorganophosphino)-naphthalenes, 1,8-(R1R2P)2C10H6 [40]. In 1a–c, the conformation of both Me2N groups is the same as that of Me2N in 2a–c, though for a different reason. A symmetrical ac arrangement of the Me groups as in 2d would cause the lone pairs to point towards each other. Rotation around the N–C(1/8) bonds by 1808 would minimize electrostatic repulsion of the lone pairs, but is obviously precluded by the steric demands of the Me groups. As a compromise, a modest rotation occurs which places the lone pairs in an antiparallel but still syn position, though at an angle with the C10 plane. In the phosphines, as a rule, the substituents R1, R2 assume positions similar to those of the Me groups in 1a-c. However, in one phosphine, R1ZPh, R2ZtBu, the groups at both P atoms reside in virtually symmetrical position with respect to the C10 skeleton. Since this arrangement is restricted to the case of most severe steric congestion, it is obvious that it is again a steric phenomenon and cannot be taken as evidence for a P/P dative bonding interaction. The sterically enforced conformation of R1, R2 would increase repulsive interaction of the lone pairs which could not be relieved by in-plane deformation. It thus becomes plausible why (exceptionally) this molecule does not resort at all to in-plane distortion (formal splay angle K0.068! [40]) and relieves steric strain exclusively by strong out-ofplane distortion which permits the lone pairs to become antiparallel at opposite faces of the C10 plane [40]. The absence of lone pair/lone pair interactions permits 2d to resort to the more common in-plane distortion. On the other hand, in protonated proton sponges, such as 1a,b$HC, 1c$2HC, the lone pair resides in the syn hemicycle as a prerequisite to engage in a N/H–NC hydrogen bond. The ’natural‘ bond lengths of very strong N/H–NC hydrogen bonds, d(N/N)Zca. 253 pm [11,23], are only slightly longer than PD so that such a hydrogen bond fits into the peri space with almost no distortion of the C10 skeleton (vide supra). On the other hand, the fact that d(N/N)OPD indicates that ca. 253 pm is about the minimum N/N distance even for strong hydrogen bonds. d(N/O)Z256.9 pm in DAN-OH [29] (i.e. again slightly longer than PD) suggests that this applies to N/H–O hydrogen bonds as well. In the case of 2d, the conformation required for the N/H–O triangular plane to coincide with the C10 plane would place the HO–Ccarb bond in the sp sector ecliptic with C10 and the C3H5 groups in the ac sectors. While sterically favourable for C3H5, it would place O at a very short distance to N, not sufficient to accommodate a N/H–O hydrogen bond, viz ca. 121 pm in the undistorted DAN system and ca. 205 pm in the distorted one actually found. As no surprise, then, HO avoids the unfavourable sp
179
conformation: C3HI5 adopts an ap position deviating by 158 from the anti-ecliptic position known to be well suited for a bulky substituent (cf. Ph in 2b, DANII in 2c), C3HII5 a sc position deviating from orthogonality by ca. 168, and O a skew syn position above face A of C10 (dihedral angle O– Ccarb–C(1)/C(8) K46.78, sc) (see Table 2). The result is again a herring bone pattern: The bonds C(C3HII5 )–Ccarb and MeB–N as well as O–Ccarb and MeA–N are not far from being parallel (deviations ca. 13 and ca. 178, respectively). The slightly larger peri space B is left to the bulkier substituent (C3HII5 ), the slightly smaller space A to the HO group. It seems, then, that again steric factors govern the arrangement. Formally, the N/O distance, 261.2 pm, would be compatible with a strong N/H–O hydrogen bond (vide supra). d(N/H)Z162 pm exceeds d(O–H)Z 103 pm by 58%. The angle N/H–O, 1628, deviates from linearity by 188. According to these criteria, the molecule would contain a N/H–O hydrogen bond which is stronger than that in 2c and which would compare not unfavourably with Steiner’s strong N/H–O hydrogen bond [34]. However, the positions of MeA,B indicate that the lone pair is far from residing in the N/H–O triangular plane. The short N/O distance may well, therefore, be conditioned by the geometric situation rather than by a hydrogen bond. Whether this arrangement results in an attractive interaction, i.e. whether it qualifies for a hydrogen bond, remains an unanswered question. At best, the energy gain by such N/H–O interaction can hardly be expected to contribute much to the energy balance of the system, and it remains to conclude that again the structure (including the conformations of the peri substituents) is the consequence of bond lengths, bond angles and skeleton geometry preserving forces in conjunction with steric repulsion forces and that a hydrogen bond plays only a subordinate, if any role. Some insight into the role of the HO group might be expected from a comparison of the DAN carbinols with analogous structures in which the prerequisites of a N/H– O hydrogen bond are lacking. The methyl ether of di(tbutyl)(naphth-1-yl)carbinol (3) is a compound analogous to 2d but with neither electronic nor steric peri interactions at all (except for those with peri-H). One of the t-butyl groups occupies a position about perpendicular to the C10 plane, whereas the other two groups share the space on the opposite side, both in skew positions, one in the syn region and the other one in the anti region. Similar to the HO group in 2a and with respect to DANII in 2c, the MeO group is capable of occupying the anti position, but the syn position (as HO in 2b and with respect to DANI in 2c) is the more stable one (dihedral angle O–Ccarb–C(1)–C(9) 228 [41]). The closely related arrangements in 3 and in 2d suggest that the preference of HO for the syn position in the latter is not conditioned by a N/H–O interaction. Replacement of Ccarb in 2 by a PC atom as a tetrahedral centre would strongly enhance the acidity of HO so that only
180
P.M. Dominiak et al. / Journal of Molecular Structure 751 (2005) 172–183
the O-deprotonated species, i.e. the phosphine oxide 4a would be stable. Unfortunately, so far no structures of DANphosphine oxides seem to have been determined. However, the data of a DAN-phosphine sulfide, 4b, are available [19]. With a chalcogen atom and two fairly large organic groups bound to the tetrahedral centre in addition to DAN, it corresponds to 2d. Unlike 2d, it gains relief from steric strain mainly by out-of-plane rather than by in-plane distortion (DANI/DANII: d(N/P)Z300.9/301.1 pm; splay angle C5.3/5.18 [10]; d(C(1)/C(8))Z252.8/256.0 pm; d(C(4)/C(5))Z247.9/246.9 pm; dihedral angles P– C(1)/C(8)–N 37.4/36.18). With respect to each of the DAN groups, as in the carbinols (except 2a) one of the organic groups resides in the ap sector; because of the diastereotopicity of DANI and DANII, this is DAN in one case and Ph in the other. Not surprisingly, the arrangement is again the herring-bone pattern (dihedral angles, DANI: MeIA–NI-C(8)I/C(1)I 89.58 (Z180–90.58), C(1)Ph-PC(1)I/C(8)I K89.28, both perpendicular; Me IB–N I– C(8)/C(1) K143.48, ac, S–P–C(1)/C(8) 37.38 (Z180– 142.78), sc; DANII: MeIIA-NII–C(8)II/C(1)II 98.98 (Z180– 81.18), ac, C(1)I–P–C(1)II/C(8)II K80.88, sc; MeIIB–NII– C(8)II/C(1)II K137.48, ac, S–P–C(1)II/C(8)II 53.08 (Z 180–1278), sc) with the chalcogen atom in sc position. As a consequence of d(P–C(1))Od(Ccarb–C(1)) and d(P–S)O d(Ccarb–O), d(N/S)Z315.4/322.5 pm exceeds d(N/O) Z 261.2 pm in 2d considerably, but is still shorter than Sr(vdW)[N, S]Z335 pm [14]. The overall similarity suggests that in the phosphine sulfide 4b and in the carbinols 2a–d the arrangement is governed by the same parameters, i.e. that in 2d, the position of HO is indeed conditioned not by a hydrogen bond but by the steric situation. An analogous structure with two equal organic groups bound to the tetrahedral centre (again PC) in addition to an 8-(s-donor) substituted naphth-1-yl group and a chalcogen atom, 4c, is provided by a recent investigation of Schmutzler et al. [42]. Two bulky peri substituents induce the molecule to resort both to strong in-plane (splay angle: C14.78!) and out-of-plane distortion (dihedral angle PC– C(1)/C(8)–P 28.98; significant non-planarity of the C10 skeleton) which permit interatomic distances d(P/P)Z 324.8 pm and d(P/Se)Z341.1 pm (both less than Sr(vdW)Z340 and 355 pm, respectively [14,43]). Even with this bulky chalcogen atom, the molecule maintains the herring-bone pattern with the hetero atom in syn position. The Ph2P substituent assumes roughly the same conformation as Me2N in the DAN compounds. Of the Ph2P(Se) substituent, one Ph is in ap position and the Se atom more pronouncedly syn than the second Ph group (dihedral angles Se–PC-C(1)/C(8) 50.98, CPh–PC–C(1)/C(8) K83.08, both sc); the sc Ph group deviates from parallelism with one of the Ph groups at the phosphino P by only ca. 108. As the NI-lone pair of 2b,c does with respect to HO, the P-lone pair points to the face of C10 on which the Se is situated (mutatis mutandis: face B): Though, in a basically
analogous structure, many parameters have been changed, the overall arrangement has remained the same. The obvious conclusions are that a) the preferred position of a chalcogen X (O, S, Se) bound to a tetrahedral peri-atom, in competition with hydrocarbon groups, is sc, regardless whether the other peri position is occupied by nitrogen, phosphorus or hydrogen, b) for this position, bond lengths and angles place X and any substituent in the other peri position, e.g. N, at a distance which is similar to d(N/X) in N/H–X hydrogen bonds, so that from the very fact of such distances no information about hydrogen bonds can be deduced, and c) in the carbinols, the energy gain of a N/H–O interaction is unable to override other factors and that the N atom can approach the H–O group only if favourable conditions are already provided. 1 H NMR provides some insight into the situation in solution. The 1H NMR spectra of 2c have previously been discussed [7]. At C238, when rotation around the C(1)-Ccarb bonds is (almost) ’free‘, but rotation around the C(8)–N bonds restricted, one (broad) singlet at dI 2.33 and one sharp singlet at dII 2.70 have been observed [7]. Similarly, 2a,b exhibited singlets at dI 2.310 and dII 2.735. Since in a series of carbinols 2, a comparable low field signal (dII 2.66–2.88) was observed only when R1ZH, aliphatic (R2Zaliphatic or aromatic), and a high field signal (dI 1.58–2.36) only when R1Zaromatic (R2 ZH, aliphatic or aromatic) [44], DdI,IIZ 0.425 can be assigned to aromatic shielding of the Me protons absorbing at dI. Correspondingly, 2d exhibited only one singlet in the low-field region, d 2.778. The two singlets of 2a/b coalesced at 387 K, from which DG‡ Z77.3 kJmol-1 is estimated [45]. This barrier is higher than but still comparable with DG‡Z46–47 kJmolK1 in 2c [7]; both figures frame the DG‡ values of (DAN)2SiR1R2, R1ZH, R2ZF (61.5), R1ZH, R2ZMeO (63.6), R1ZMe, R2ZPh (57.0 kJmolK1) [46] which, in our opinion, likewise are due to restricted rotation (rather than N/Si dative bonding interactions) [7]. Consequently, no influence of the N/H– O interactions in the carbinols upon DG‡ is discernible. In the secondary carbinols 2a–c, the low-field absorption of the HO proton (2a/b: d7.76, 2c: d7.96) presumably indicates hydrogen bonding. The HO proton of 2d (as of other tertiary DAN-carbinols [47]) absorbs at even much lower field (d 11.05) so that again hydrogen bonds are likely. However, since intermolecular hydrogen bonds are possible, no conclusions concerning intramolecular N/H–O hydrogen bonds can be drawn [48].
3. Conclusion As the uniform result of our investigation, it emerged that in carbinols 2, N/H–O close contacts are observed only if the geometric situation is favourable, but that in the interplay of bond lengths, bond angles, skeleton deformations, steric repulsions and hydrogen bonds, the latter do not play a decisive role. By analogy, we expect that other
P.M. Dominiak et al. / Journal of Molecular Structure 751 (2005) 172–183
attractive forces which might contribute to the overall energy of a DAN-X molecule to about the same degree or less, would likewise be unable to manifest themselves in the investigated parameters. It seems, then, unlikely that information concerning weak dative interactions between peri substituents D and A (D/A interactions) can be obtained by structural investigations of 1-A-8-Dnaphthalenes.
4. Experimental The synthesis of bis(8-dimethylamino-naphth-1-yl) carbinol (2c) has been described previously [7]. 4.1. (R,S)-(8-Dimethylamino-naphth-1-yl)(phenyl)carbinol (2a/b) 8-Dimethylamino-naphth-1-yl lithium was obtained from 10.0 ml (61 mmol) 1-dimethylamino-naphthalene and 54 ml (86 mmol) of a 1.6 M solution of n-butyl lithium in n-hexane, as described previously [5]. 8.7 ml (86 mmol) benzaldehyde were added dropwise. When a transient precipitate had dissolved, 10 ml of water and 60 ml of 2N HCl were added. The aqueous layer was made alkaline and extracted with diethyl ether. Evaporation of the organic layer and the ether extracts yielded the carbinol which was recrystallized from ethanol; yield 11.3 g (67%, m. p. 107– 1088C.—1H NMR (300 MHz, in CDCl3): d 2.310, s, 3 H; 2.735, s, 3 H (2 CH3); 6.347, d, 3JZ7.7 Hz, 1 H (Ccarb–H); 7.19–7.45, m, 9 H (C6H5, 4 DAN-H), 7.75–7.80, m, 3 H (2 DAN-H, OH at 7.757, d, 3JZ7.7 Hz); (300 MHz, in C2D2Cl 4, 1148C): d 2.535, s (N(CH 3) 2). 13C NMR (75.5 MHz, in CDCl3): d 46.30, q, 47.72, q, 76.37, d, 120.45, d, 124.97, d, 125.30, d, 126.52, d, 126.88, d, 127.85, d, 127.88, d, 128.74, s, 129.09, d, 129.40, d, 136.87, s, 139.27, s, 145.06, s, 150.03, s.—C19H19NO (277.4): calc. C 82.28, H 6.90, N 5.05, found C 82.02, H 6.92, N 5.10. 4.2. Di(cyclopropyl)(8-dimethylamino-naphth-1-yl)carbinol (2d) To a suspension of 8-dimethylamino-naphth-1-yl lithium prepared (as before) from 1.0 ml (6.1 mmol) 1-dimethylamino-naphthalene and 5.4 ml (8.6 mmol) of a 1.6 M solution of n-butyl lithium in n-hexane, 0.67 g (6.09 mmol) di(cyclopropyl)ketone were added at K788C. Within 24 h, the mixture was allowed to warm up to room temp., then poured on ice and treated with 20 ml of aqueous ammonium chloride. The aqueous layer was extracted with diethyl ether. Evaporation of the organic layer and the ether extracts yielded an oil which was chromatographed on silica with t-butyl-methyl ether/n-hexane (1:4). The carbinol was eluted with Rf Z0.20. Recrystallization from petrol ether (b. p. 60–908C) yielded 1.17 g (68%), m. p. 67–69 8C. 1H NMR (200 MHz, in CDCl3): d ca. 0.52, m, 8H (CH2); 1.31, m, 2H
181
(CH); 2.778, s, 6H (N(CH3)2); 7.37–8.11, m, 6H (DAN-H); 11.054, s, 1H (OH). 13C NMR (50 MHz, in CDCl3): d 1.34, t, 3.29, t, 22.58, d, 47.11, q, 120.16, 124.73, 124.83, 127.40, 128.47, 128.89, 6 d, 129.34, 137.10, 145.95, 149.51, 4 s. C19H23NO (281.4): calc. C 81.10, H 8.24, N 4.98, found C 80.64, H 8.52, N 5.12. 4.3. X-ray diffraction All measurements of diffractional data were performed on a KUMA KM4CCD k-axis diffractometer with graphitemonochromated Mo Ka radiation. The crystals were positioned at 65 mm from the KM4CCD camera. In the case of 2a/b, ca. 612 frames were measured at 0.758 intervals with a counting time of 30 s for each frame, whereas, for 2d, 768 framed at 0.68 with 40 s per frame were collected. The data were corrected for Lorentz and polarization effects. No absorption correction was applied. Data reduction and analysis were carried out with the Kuma Diffraction (Wrocław, Poland) programs. The structures were solved by direct methods [49] and refined using SHELXL [50]. The refinement was based on F2 for all reflections except those with very negative F2. Weighted R factors wR and all goodness-of-fit S values are based on F2. Conventional R factors are based on F with F set to zero for negative F2. The F2oO2s(F2o) criterion was used only for calculating R factors and is not relevant to the choice of reflections for the refinement. The R factors based on F2 are about twice as large as those based on F. All hydrogen atoms were located from a differential map and refined isotropically. Scattering factors were taken from Tables 6.1.1.4 and 4.2.4.2 in Ref. [51]. Selected most important crystallographic data are shown in Table 1. Full crystallographic data (excluding structural factors) for the structures reported in this paper have been deposited with the Cambridge Crystallographic Data Centre and allocated the deposition numbers: CCDC 253665 and CCDC 253666 for 2a/b and 2d, respectively. Copies of the data can be obtained free of charge on application to CCDC, 12 Union Road, Cambridge CB2 1EW, UK (Fax: Int. code C (1223)336-033; E-mail:
[email protected]).
Acknowledgements Financial support by the Volkswagen Foundation (Hannover) (project Experimental and theoretical conformational analysis of organic compounds in solution) is gratefully acknowledged. PMD and KW are grateful for the Polish KBN PhD research grant No. 4 T09A 121 25. The X-ray measurements were undertaken in the Crystallographic Unit of the Physical Chemistry Laboratory at the Chemistry Department of the University of Warsaw.
182
P.M. Dominiak et al. / Journal of Molecular Structure 751 (2005) 172–183
References [1] Part 12: G.P. Schiemenz, Z. Naturforsch. 59b (2004) 807–816. [2] G.P. Schiemenz, R. Bukowski, L. Eckholtz, B. Varnsku¨hler, Z. Naturforsch, 55b (2000) 12–20. [3] G.P. Schiemenz, Z. Anorg, Allg. Chem. 628 (2002) 2597–2604. [4] Cf.,e.g. K. Mu¨ller-Dethlefs, P. Hobza, Chem. Rev. 100 (2000) 143– 167. [5] G.P. Schiemenz, S. Po¨rksen, P.M. Dominiak, K. Woz´niak, Z. Naturforsch 57b (2002) 8–18. [6] G.P. Schiemenz, C. Na¨ther, S. Po¨rksen, Z. Naturforsch 58b (2003) 663–671. [7] G.P. Schiemenz, B. Schiemenz, S. Petersen, C. Wolff, Chirality 10 (1998) 180–189. [8] G.P. Schiemenz, Chem. Listy 92 (1998) 269–270. [9] G.P. Schiemenz, Phosphorus, Sulfur, Silicon Relat. Elem. 163 (2000) 185–202; G. Dyker, M. Hagel, G. Henkel, M. Ko¨ckerling, C. Na¨ther, S. Petersen, G.P. Schiemenz, Z. Naturforsch 56b (2001) 1109–1116; G.P. Schiemenz, C. Na¨ther, S. Po¨rksen, Z. Naturforsch 58b (2003) 59– 73. [10] G.P. Schiemenz, S. Po¨rksen, C. Na¨ther, Z. Naturforsch 55b (2000) 841–854. [11] G.P. Schiemenz, C. Na¨ther, Z. Naturforsch 57b (2002) 309–318. [12] G.P. Schiemenz, S. Petersen, S. Po¨rksen, Z. Naturforsch. 58b (2003) 715–724. Erratum: On p. 721, right column, line 3, replace [2] by [5]. [13] P. Kilian, H.L. Milton, A.M.Z. Slawin, J.D. Woollins, Inorg. Chem. 43 (2004) 2252–2260. [14] S.S. Batsanov, Izvest. Akad. Nauk, Ser. Khim. (1995) 24–29; S.S. Batsanov, Russian Chem. Bull. 44 (1995) 18–23. [15] C. Brelie`re, F. Carre´, R.J.P. Corriu, M. Poirier, G. Royo, J. Zwecker, Organometallics 8 (1989) 1831–1833; F. Carre´, G. Cerveau, C. Chuit, R.J.P. Corriu, N.K. Nayyar, C. Re´ye´, Organometallics 9 (1989) 1989– 1991; F. Carre´, G. Cerveau, C. Chuit, R.J.P. Corriu, C. Re´ye´, Angew. Chem. 101 (1989) 474–476; F. Carre´, G. Cerveau, C. Chuit, R.J.P. Corriu, C. Re´ye´, Angew. Chem. Int. Ed. Engl. 28 (1989) 489; F. Carre´, G. Cerveau, C. Chuit, R.J.P. Corriu, C. Re´ye´, N. J. Chem. (Nouv. J. Chim.) 16 (1992) 63–69; K. Boyer-Elma, F.H. Carre´, R.J.P. Corriu, W.E. Douglas, J. Chem. Soc., Chem. Commun. (1995) 725– 726; K. Tamao, K. Nagata, M. Asahara, A. Kawachi, Y. Ito, M. Shiro, J. Am. Chem. 117 (1995) 11592–11593; K. Tamao, M. Asahara, A. Kawachi, J. Organomet. Chem 521 (1996) 325–334; K. Tamao, M. Asahara, Guang-Ri Sun, A. Kawachi, J. Organomet. Chem. 574 (1999) 193–205; K. Tamao, M. Asahara, A. Toshimitsu, Chem. Lett. (2000) 658–659; K. Tamao, M. Asahara, T. Saeki, Shengyu Feng, A. Kawachi, A. Toshimitsu, Chem. Lett. (2000) 660–661; M. Spiniello, J.M. White, Organometallics 19 (2000) 1350–1354; K. Tamao, M. Asahara, A. Kawachi, A. Toshimitsu, J. Organomet. P Chem. 643–644 (2002) 479–486; 300 pm!d(N/Si) ! r(vdW)[N, Si]: (n) K. Tamao, Y. Tarao, Y. Nakagawa, K. Nagata, Y. Ito, J. Organomet. Chem. 12 (1993) 1113–1120; K. Tamao, M. Asahara, T. Saeki, A. Toshimitsu, Chem. Lett. (1999) 335–336; M.J. Hanson, M. Spiniello, J.M. White, J. Organomet. Chem. 654 (2002) 3–7. [16] C. Brelie`re, F. Carre´, R.J.P. Corriu, G. Royo, M. Wong Chi Man, J. Lapasset, Organometallics 13 (1994) 307–314; K. Tamao, M. Asahara, T. Saeki, A. Toshimitsu, Angew. Chem. 111 (1999) 3520–3523; K. Tamao, M. Asahara, T. Saeki, A. Toshimitsu, Angew. Chem. Int. Ed. 38 (1999) 3316–3318. [17] H.W.W. Ehrlich, Acta Crystallogr. 10 (1957) 699–705. [18] C. Chuit, R.J.P. Corriu, P. Monforte, C. Reye´, J.-P. Declercq, A. Dubourg, Angew. Chem., 105 (1993) 1529–1531; C. Chuit, R.J.P. Corriu, P. Monforte, C. Reye´, J.-P. Declercq, A. Dubourg, Angew. Chem. Int. Ed. Engl., 32 (1993) 1430; F. Carre´, C. Chuit, R.J.P. Corriu, A. Mehdi, C. Reye´, Inorg. Chim. Acta, 250 (1996) 21– 28; C. Chuit, R.J.P. Corriu, C. Reye´, in: K.-y. Akiba (Ed.), Chemistry of Hypervalent Compounds, Wiley, New York, 1999, pp. 81–146.
[19] M. Chauhan, C. Chuit, R.J.P. Corriu, C. Reye´, J.-P. Declercq, A. Dubourg, J. Organomet. Chem. 510 (1996) 173–179. [20] C. Chuit, C. Reye´, Eur. J. Inorg. Chem. (1998) 1847–1857. [21] R.O. Day, T.K. Prakasha, R.R. Holmes, H. Eckert, Organometallics 13 (1994) 1285–1293 (p. 1290); T.K. Prakasha, S. Srinivasan, A. Chandrasekaran, R.O. Day, R.R. Holmes, J. Am. Chem. Soc. 117 (1995) 10003–10009 (p. 10006). [22] N.V. Timosheva, A. Chandrasekaran, R.O. Day, R.R. Holmes, Inorg. Chem. 37 (1998) 4945–4952. [23] R.W. Alder, A.G. Orpen, R.B. Sessions, J. Chem. Soc. Chem. Commun. (1983) 999–1000; H. Bock, T. Vaupel, C. Na¨ther, K. Ruppert, Z. Havlas, Angew. Chem. 104 (1992) 348–350; H. Bock, T. Vaupel, C. Na¨ther, K. Ruppert, Z. Havlas, Angew. Chem. Int. Ed. Engl. 31 (1992) 299–301; H. Bock, T. Vaupel, H. Scho¨del, J. Prakt. Chem. 339 (1997) 26–37. [24] Z. Malarski, T. Lis, E. Grech, J. Nowicka-Scheibe, K. Majewska, J. Mol. Struct. 221 (1990) 227–238; K. Wozniak, T.M. Krygowski, B. Kariuki, W. Jones, E. Grech, J. Mol. Struct. 240 (1990) 111–118. [25] H.A. Staab, C. Krieger, G. Hieber, K. Oberdorf, Angew. Chem. 109 (1997) 1946–1949; H.A. Staab, C. Krieger, K. Hieber, K. Oberdorf, Angew. Chem. Int. Ed. Engl. 36 (1997) 1884–1886. [26] The shortest N/H–NC hydrogen bond reported so far, d(N/N) Z 246.4 pm, has exactly the same length as PD: R. Destro, T. Pilati, M. Simonetta, E. Vogel, J. Am. Chem. Soc. 107 (1985) 3185–3191. [27] T. Steiner, Chem. Commun. (1999) 313–314. [28] W. Klyne, V. Prelog, Experientia, 16 (1960) 521–523; J. Rigaudy, S.P. Klesney, International Union of Pure and Applied Chemistry, Organic Chemistry Division, Commission on Nomenclature of Organic Chemistry: Nomenclature of Organic Chemistry, Sections A, B, C, D, E, F and H, Pergamon Press, Oxford, 1982, p. 484. Cf. S. Kane, W.H. Hersh, J. Chem. Educ., 77 (2000) 1366. [29] E. Grech, J. Nowicka-Scheibe, Z. Olejnik, T. Lis, Z. Pawelka, Z. Malarski, L. Sobczyk, J. Chem. Soc., Perkin Trans. II (1996) 343– 348. [30] A. Toshimitsu, T. Saeki, K. Tamao, Chem. Lett. (2002) 278–279. [31] T. Herbst, G.P. Schiemenz, C. Wolff, Magn. Reson. Chem. 26 (1988) 608–611; T. Herbst, G.P. Schiemenz, Z. Naturforsch 44b (1989) 866– 868. [32] W.B. Schweizer, G. Procter, M. Kaftory, J.D. Dunitz, Helv. Chim. Acta 61 (1978) 2783–2808; J. Clayden, C. McCarthy, M. Helliwell, Chem. Commun. (1999) 2059–2060. [33] Cf. D. Casarini, L. Lunazzi, A. Mazzanti, J. Org. Chem. 62 (1997) 3315–3323. [34] T. Steiner, C.C. Wilson, I. Majerz, Chem. Commun. (2000) 1231– 1232. [35] Cf. T. Steiner, Chem. Commun. (1999) 2299–2300. [36] V.R. Pedireddi, G.R. Desiraju, J. Chem. Soc., Chem. Commun. (1992) 988–990; G.R. Desiraju, Angew. Chem. 107 (1995) 2541– 2558; G.R. Desiraju, Angew. Chem. Int. Ed. Engl. 34 (1995) 2311–2327. [37] V.R. Thalladi, A. Gehrke, R. Boese, New J. Chem. (2000) 463–470. [38] C. Chuit, R.J.P. Corriu, P. Monforte, C. Reye´, J.-P. Declercq, A. Dubourg, J. Organomet. Chem. 511 (1996) 171–175. [39] T. Barth, C. Krieger, F.A. Neugebauer, H.A. Staab, Angew. Chem. 103 (1991) 1006–1008; T. Barth, C. Krieger, F.A. Neugebauer, H.A. Staab, Angew. Chem. Int. Ed. Engl. 30 (1991) 1028; K. Wozniak, H. He, J. Klinowski, B. Nogaj, D. Lemanski, D.E. Hibbs, M.B. Hursthouse, S.T. Howard, J. Chem. Soc., Faraday Trans. 91 (1995) 3925–3932. [40] A. Karac¸ar, H. Tho¨nnessen, P.G. Jones, R. Bartsch, R. Schmutzler, Heteroatom Chem. 8 (1997) 539–550. [41] D. Casarini, L. Lunazzi, A. Mazzanti, E. Foresti, J. Org. Chem. 63 (1998) 4746–4754. [42] A. Karac¸ar, M. Freytag, H. Tho¨nnessen, J. Omelanczuk, P.G. Jones, R. Bartsch, R. Schmutzler, Z. Anorg, Allg. Chem. 626 (2000) 2361– 2372.
P.M. Dominiak et al. / Journal of Molecular Structure 751 (2005) 172–183 [43] Schmutzler et al. used older r(vdW) data and thus arrived at much more impressive differences. Besides other arguments, the poor precision of r(vdW) values invalidates Holmes’s procedure of calculating degrees of hypercoordination [21]. [44] S. Petersen, Diploma thesis, University of Kiel, 1993. [45] M. Hesse, H. Meier, B. Zeeh, Spektroskopische Methoden in der organischen Chemie, 4th ed., Thieme, Stuttgart, 1991, p. 96. [46] C. Brelie`re, R.J.P. Corriu, G. Rojo, J. Zwecker, Organometallics 8 (1989) 1834–1836. [47] S. Petersen, PhD thesis, University of Kiel 1999.
183
[48] Note that on the one hand, the X-ray structure determination revealed an intermolecular O/H–O bond in crystalline 2a/b and that, on the other hand, in the absence of 8-Me2N, the carbinols C10H7-C(OH)R2, RZ Et, iPr, in solution do not form such bonds (d(H–O) 1.97, 1.82) [33]. [49] G.M. Sheldrick, Acta Crystallogr. A 46 (1990) 467–473. [50] G.M. Sheldrick, SHELXL97. Program for the Refinement of Crystal Structures, University of Go¨ttingen, Germany. [51] A.J.C. Wilson, International Tables for Crystallography, vol. C, Kluwer, Dordrecht, 1992.
