摘要:

Mass spectrometric approaches to the reactivity of transient neutrals Christoph A. Schalley Georg Hornung Detlef Schr�oder and Helmut Schwarz* Institut f�ur Organische Chemie Technische Universit�at Stra ße des 17. Juni 135 10623 Berlin Germany During the past few years Neutralisation–Reionisation Mass Spectrometry (NRMS) has developed from a method for the generation and structural characterisation of elusive and highly reactive neutral molecules to a useful tool for probing their chemical reactivity. Three major principles can be distinguished (i) peak shape analysis (ii) activation of the neutrals by collisions or light and (iii) variation of the neutrals’ lifetimes. Several methodological approaches are discussed in conjunction with illustrating examples for the chemical reactivity of transient neutrals.1 Introduction The use of mass spectrometers for the examination of the unimolecular chemistry of neutral molecules seems at first sight as odd as squaring the circle because magnetic and electrostatic fields primarily afford mass analysis of ionic species. However the discovery that one-electron transfer to or from fast moving ion beams can be achieved by high energy collisions led to the development of neutralisation–reionisation mass spectrometry (NRMS) which was first applied to chemical problems almost 20 years ago. NRMS has proved to be of enormous value for the formation of elusive and highly reactive neutrals which are difficult to generate by other means due to their propensity toward unimolecular fragmentation or fast bimolecular rear- Christoph A.Schalley born 1968 in Krefeld Germany studied chemistry in Freiburg and Berlin as a fellow of the Studienstiftung des Deutschen Volkes. He obtained his PhD with Professor Schwarz in 1997. Georg Hornung born 1966 in Zalaegerszeg Hungary studied chemistry in Berlin and Canterbury and is currently working on his PhD with Professor Schwarz. Detlef Schr�oder born 1963 in Wilster Germany. He obtained his PhD with Professor Schwarz at the TU Berlin in 1992 and is currently working as research assistant in the laboratory of Professor Schwarz. He is the recipient of the 1993 Schering Award and the 1997 Mattauch-Herzog Preis. Christoph A. Schalley Detlef Schr�oder Georg Hornung rangements.The literature up to 1994 dealing with NRMS has been documented in several reviews.1–8 The generation of acepentalene 1 (Scheme 1),9 ylides such as 2 and 3,10 carbenes HO 1 2 3 (X = NH S) O OH H OH 4 Helmut Schwarz NH HN NH22 H NH2 7 6 91 5 Scheme 1 like 4 and 6 which in the gas phase do not isomerize to 5 and 7,11,12 and several cumulenes13 of broad interest represent some more recent and instructive examples for the performance of Helmut Schwarz born in 1943 spent four years as a chemical technician before reading chemistry at the Technische Universit �at Berlin (TUB) where he obtained his PhD under the supervision of F. Bohlmann in 1972 and his habilitation in 1974. Since 1978 he has been at the TUB as a Professor of Chemistry and has resisted all temptations to leave Berlin for good.He has held visiting appointments in Cambridge Lausanne Jerusalem Innsbruck Haifa Paris Auckland and Canberra. Among his several awards are the Otto-Klung Otto- Bayer and the Leibniz Prizes and the Max-Planck Research Award which he received jointly with C. Lifshitz. In 1992 the Hebrew University of Jerusalem conferred an honorary doctorate on him and he received in 1994 the J. J. Thompson Gold Medal. Dr Schwarz will be the 1998 recipient of the Liebig Medal of the German Chemical Society. Chemical Society Reviews 1998 volume 27 H H N N X H H N N X this method and its application to different fields of chemistry. Let us first briefly resume the key features of NRMS.In a neutralisation–reionisation experiment [Fig. 1(a)] a massselected fast-moving ion beam is subjected to two sequential collision events with appropriate target gases in gas cells along the flight path. In the first collision a fraction of the ions is neutralised by a single electron transfer. Then the remaining ionic species are deflected by applying a high voltage to a deflector electrode which is located between the two gas cells. Subsequently the fast moving neutrals are reionised by yet another collision in the second gas cell. Finally the reionised particles are analysed by standard mass spectrometric means and serve as an indirect probe for the existence of the neutral intermediates. Two particular features of NR mass spectra have attracted considerable attention.(i) The detection of a recovery signal at the m/z ratio of the original projectile ions indicates that these ions have survived the whole series of neutralisation and reionisation events. Accordingly the corresponding neutral counterparts have lifetimes exceeding the time delay between the two electron transfer steps and are typically covering the microsecond regime. (ii) The fragmentation pattern of NR spectra often contains valuable structural information about the connectivities of the neutral species as well as the projectile and survivor ions. The NR approach which is complementary to other experimental techniques that are employed for the investigation of isolated neutral molecules such as matrix isolation or molecular beam spectroscopy offers three features (i) With the exception of pulsed ionisation methods the whole repertoire of ionisation Fig.1 Typical experimental set-ups for variants of neutalisation-reionisation (NR) experiments (a) conventional NR procedure with mass selected AB+/2 (b) collisional activation of survivor ions AB+/2 generated in a NR experiment (NR/CA) (c) collision-induced dissociative ionisation (CIDI) of neutrals N1 and N2 originating from metastable transitions of the parent ions AB+/2 and (d) neutral fragment reionisation (NfR) of neutral N1 and N2 generated by collisional activation of the parent ions AB2/+. Chemical Society Reviews 1998 volume 27 92 methods can be applied for the formation of the precursor ions for NRMS experiments.The ion connectivities can well be identified by collisional activation (CA) experiments. Since ion structures different from the connectivities of conventional neutral molecules are often easily accessible NRMS allows for the generation of unusual neutral species. (ii) In keV collisions the electron transfer proceeds within ca. 10215 s and can therefore be regarded as vertical i.e. geometry changes during the neutralisation and reionisation steps are negligible. Accordingly Franck–Condon factors14 associated with neutralisation and reionisation dominate the efficiency of the NR process the internal energy deposited in the neutral and reionised species as well as the fragmentation pattern (Fig. 2). (iii) Solvent effects and other intermolecular interactions are strictly excluded in the high vacuum existing inside a mass spectrometer.NR mass spectrometric techniques can be performed in several variants which contribute significantly to the large body of NR literature.1–8 (i) Depending on the charges of the projectile and recovery ions the NR procedure as depicted in Fig. 1(a) can be performed in four variants denoted according to McMahon et al.4 as 2NR2 2NR+ +NR2 and +NR+. (ii) For structural characterisation or reactivity studies the survivor ions can be mass selected with the following sector and submitted to collisional activation in the subsequent field-free region provided that its intensity is large enough [NR/CA; Fig. 1(b)].15 Comparison of the NR/CA mass spectra with conventional CA spectra recorded under the same conditions provides information about the geometric structure of the reionised ions and the purity of the parent ion beam.(iii) Neutrals generated in unimolecular or collision-induced decompositions can be Fig. 2 (a) Schematic +NR+ spectrum of AB+ being geometrically similar to neutral AB. This results in favourable Franck–Condon factors and hence little fragmentation. (b) Schematic +NR+ spectrum of AB+ for whion and neutral geometries largely differ and thus Franck–Condon factors become unfavourable. AB is formed with a higher internal energy (dotted line) and upon reionisation AB+ may be formed rovibrationally excited above the dissociation threshold. Thus although both AB+ and AB are stable species the recovery signal may be small or even absent and fragmentation prevails.characterised by collision-induced dissociative ionisation [CIDI; Fig. 1(c)] and neutral fragment reionisation [NfR; Fig. 1(d)] respectively.16,17 Except for a few investigations up to 1994 the applications of the NR methodology were focussed on the unique capability of probing the existence and structure of unconventional neutral molecules and as a spin-off to elucidate cation structures for which differentiation between isomers turned out to be difficult by collisional activation. Due to some major obstacles the question of the neutrals’ chemical reactivity however has only scarcely been addressed. For example the unambiguous analysis of the neutrals’ reactivity based on the analysis of mass spectra is often difficult to achieve due to a superposition of the fragmentations of the neutral intermediate the projectiles and the reionised ions in the NR mass spectra.Further the microsecond timescale of NR experiments often renders the detection of neutral decompositions difficult due to their low intensities. In addition most of the methods discussed below require an advanced experimental set-up for an unequivocal detection of the neutrals’ fragmentations. Nevertheless it remains challenging to study the chemical reactivity of transient neutrals produced by NR mass spectrometry because many of these species have unique electronic or structural features and cannot be generated by other experimental methods.In the past few years several modifications of the conventional NR technology have been developed in order to extend its capability toward neutral reactivity. Far from giving a comprehensive overview we will discuss these methods in the following sections in an approximately chronological order together with some instructive examples. In general three different approaches can be distinguished. (i) Peak shape analysis may provide information about the kinetic energy releases associated with decomposition reactions of ions and/or neutrals. In favourable cases the energy releases are different for ions and neutrals which then is reflected in the peak shapes. (ii) Fragmentations of the neutrals species can be induced by collisions with an appropriate target gas or by photoexcitation with a laser beam.This approach does not however monitor the intrinsic unimolecular reactivity of the transient neutrals. (iii) Variations of the neutrals’ lifetimes by different lengths of the flight paths between neutralisation and reionisation allow for an analysis of the unimolecular processes. The advantages and drawbacks of each method will be addressed in the following sections. 2 Early studies In the early days of NR mass spectrometry Gellene and Porter1 derived some information about neutral species from the peak shape of the neutral beam. Upon fragmentation of the hot neutral transients a fraction of the internal energy is released as kinetic energy thus resulting in line broadening of the peak.This information is also contained in the peak shapes of conventional NR mass spectra and composite peaks may indicate superpositions of neutral and ionic decompositions leading to the same products with different kinetic energy releases. Los et al. developed the differential translational spectroscopy18 which allows the study of the unimolecular decay of neutral molecules generated in a collision of a fast moving ion with a target gas. By this approach the mass ratio of both neutral fragments and the amount of kinetic energy released can be measured. Neutral CH3CO. radicals for example reveal dissociation into CH3 . and CO after randomisation of the internal energy. The measured kinetic energy releases even provide information about the dynamic properties of the reaction in that a model for the conversion of bending motion into fragment translation has been derived.Neutral CH3CO. radicals have also been examined by CIDI of CH3CO. generated unimolecularly from bisacetyl and by NR of CH3CO+ as formed from ionised acetone.19 In the CIDI mass spectrum ionised acetyl radical represents the base peak while the recovery signal is weak in the NR mass spectrum. By peakshape analysis it can be shown that CH3CO. undergoes two fragmentations [eqns. (1) and (2)] when formed by neutralisation of CH3CO+. (1) (2) CH CH3CO.?CH2CO + H. 3CO.?CO + CH3 . These observations were explained by Frank–Condon factor dominated neutralisation of the linear CH3CO+ cation giving rise to excited CH3CO.radicals which rapidly dissociate. In contrast unimolecular dissociation of bisacetyl leads to the bent ground state of the CH3CO. radical which hardly dissociates. Another approach to probe the neutrals’ reactivity was introduced in 1986 when McLafferty and co-workers examined 93 Chemical Society Reviews 1998 volume 27 ylides such as CH2XH (X = F Cl OH NH2) and their conventional counterparts CH3X by NR experiments applying Hg and He as target gases for neutralisation and reionisation respectively.20 By varying the helium pressure within the reionisation cell they found significant changes in the NR mass spectra which were attributed to collision-induced dissociation reactions of the neutral intermediate. For example in CH3Cl+.the weakest bond is a C–H bond while in neutral CH3Cl the C–Cl bond is cleaved more readily. By increasing the reionisation gas pressure that results in a decrease of transmittance (T) from 90% to 30% the Cl+ signal in the NR mass spectrum of CH attributed to dissociation of neutral CH3Cl into CH3 . + Cl. 3Cl increased dramatically. This has been induced by additional collisions before reionisation. In contrast a drastic increase of the signals for Cl+ and HCl+. was observed for the ylide isomer CH2ClH. Unfortunately this technique20 suffers from some ambiguities concerning the sequence of collision induced decomposition and reionisation. It is not quite clear whether reionisation comes first followed by a decay of the reionised ions or vice versa.In short this method can only be applied to rather small systems in which distinctly different dissociation channels of neutral and ionic species exist. Thermochemical data or theoretical calculations may then help to decide whether the fragments are due to neutrals or ions. For larger molecules however it becomes increasingly more difficult to analyse their reactivity merely on the basis of NR mass spectra obtained with different reionisation gas pressures. Furthermore thermochemical data are often incomplete for these species and therefore cannot help to decide whether neutral or ionic processes are monitored. 3 Collisional activation of neutral species the NCR method In order to avoid these uncertainties a modified experimental set-up was developed by McLafferty and others15 which allows for neutralisation collisional activation of the neutrals and reionisation (NCR) in three separate and differentially pumped collision cells (Fig.3). The collision conditions in the second cell can be varied independently with respect to the choice of appropriate gases and transmittances. For collisional activation helium is most often used since it reduces the contribution of electron transfer processes. In these experiments two deflector electrodes are required in front of and behind the second collision cell for separating the neutral beam after neutralisation and deflecting all ions formed in the second cell. Using the NCR approach Turecek et al. were able to demonstrate that neutral cyclohexa-2,4-dienone21 8 (Scheme 2) OH O 8 Scheme 2 9 does not readily isomerise to the more stable phenol molecule 9.Even upon collisional activation of neutral 8 in +NCR+ experiments some distinct differences between the +NCR+ Fig. 3 Set-up of a NCR experiment. After neutralisation and deflection of the remaining ions the neutral beam is subjected to collisions in a second collision cell and subsequently reionised in the third one. A second deflector electrode ensures that neutrals reionised in the second collision cell are not monitored. Chemical Society Reviews 1998 volume 27 94 spectra of 8 and 9 remained. The results were interpreted in terms of symmetry-forbidden [1,3]- and [1,7]-sigmatropic hydrogen shifts. The NR and NCR mass spectra of the [C2 H4 O]+.isomers 10+.–16+. have provided a comprehensive set of experimental data which allows for the construction of the neutral semiquantitative [C2,H4,O] potential-energy surface as depicted in Fig. 4.22 The three most stable isomers i.e. vinyl alcohol 10 acetaldehyde 12 and oxirane 14 correspond to conventional structures while the biradicals 13 and 15 and the carbenes 11 and 16 are much higher in energy. NCR experiments were used to determine the extent of isomerisations between these local minima and to derive a qualitative estimate of the relative barrier heights. In this study it was found that 11 as well as 13 are both stable but upon collisional activation predominant isomerisation to acetaldehyde 12 was observed.Furthermore TS11/12 and TS12/13 are predicted to be lower in energy than TS10/11 and TS13/14 respectively. A path for the direct 10? 12 rearrangement involving a 1,3-hydrogen shift was not found. Similarly it could be established that TS15/16 must be higher than the exit channel of lowest energy (CH3 . + HCO.) because methoxy carbene 16 does not rearrange to any of the other [C2 H4 O] isomers. In conclusion this early example demonstrates the performance of NRMS to examine neutrals including some unusual species such as biradicals and carbenes. For NCR studies it is quite advantageous to perform a series of experiments at several different gas pressures for collisional activation of the neutral beam. As an example let us refer to a recent study dealing with the unimolecular decay of methyl hydroperoxide cation radicals CH3OOH+23 In order to obtain information about the structure and stability of ionised CH3OOH+.+NR+ and +NCR+ experiments (Fig. 5) have been performed which—besides intense recovery signals—revealed fragments such as O2 +. HO2 + and CH2OOH+. These structureindicative fragment ions unambiguously point towards a peroxidic connectivity for the cation as well as its neutral counterpart. In addition strong support for this structural assignment came from the observation of an O–O bond cleavage of the neutral species. While in neutral CH3OOH the O–O bond is weak in the corresponding cation radical the C–O bond is cleaved more readily due to changes in bond strengths. These theoretically predicted features are indeed reflected in the NR and NCR mass spectra which contain fragments indicative of these processes.For example CH3 + and HO2 + are generated by C–O bond rupture while HO+ and HCO+ indicate O–O bond cleavage. The latter fragment ion is formed from the quite unstable CH3O+ by H2 loss. Comparing the spectra displayed in Figs. 5(b–d) it becomes obvious (Table 1) that the CH3 + :HO2 + and HO+ :HCO+ ratios remain approximately constant at 2 1 and 1 11 irrespective of the helium pressure used for collisional activation of neutral CH3OOH. In contrast with decreasing transmittance the HO+ and HCO+ intensities increase relative to those of CH3 + and HO2 + indicating O–O bond cleavage as a major process at the neutral stage.Thus the structural assignment of a peroxidic connectivity cannot only be based on characteristic fragments but also on the observation of the expected reactivity of the proposed neutral. Fig. 4 [C2,H4,O] potential-energy surface derived from several NR variants with neutralised [C2,H4,O]+ isomers 10–16. Heats of formation (DHf in kcal mol21) have been taken from ref. 22. + helium pressure (Xe 80% T; He 60% T; O Fig. 5 (a) Conventional (SD)-+NR+ mass spectrum (Xe 80% T; O2 70% T) of CH3OOH+. using the second and third of three collision cells. (b) (LD)- NR+ spectrum (Xe 80% T; O2 70% T) with an increased flight path of the neutral intermediates. (c) Same spectrum as (b) applying helium in the second cell for collisional activation of the neutral (+NCR+; Xe 80% T; He 80% T; O2 70% T).(d) +NCR+ mass spectrum equal to (c) with increased 2 70% T). Table 1 Ratios of intensities O–O versus C–O bond dissociation products in the long distance +NR+ +NCR+ (He 80% T) and +NCR+ (He 60% T) mass spectra of CH3OOH+. [Figs. 5(b,c,d)]a Ion pair NCR (60% T) NCR (80% T) LD-NRb (100% T) CH3 + :HO2 + HO+ :HCO+ CH3 + :HO+ HO2 + :HO+ CH3 + :HCO+ 2.0 1.0 1.0 11.1 1.4 1.0 1.0 1.3 1.0 7.1 1.0 13.2 2.0 1.0 1.0 11.3 2.0 1.0 1.0 1.0 1.0 5.6 1.0 11.1 HO2 + :HCO+ a 1.0 1.0 1.0 10.2 1.2 1.0 1.0 1.4 1.0 7.3 1.0 12.5 The conventional +NR+ spectrum [Fig. 5(a)] is not included in the Table because the relative intensities may be subject to mass discrimination (see text).b Long distance-NR mass spectrum (see Chapter 4). A drawback of NCR is that collision induced and not unimolecular fragmentations are monitored. While collisional activation is advantageous for a structural characterisation due to direct bond ruptures it does not necessarily give insight into low-energy pathways which often involve rearrangements. Indeed a method for the investigation of the decay of metastable neutral species is provided by variation of the flight path and as a consequence the lifetimes of the neutrals. A longer flight path is expected to result in increased intensities for those processes which take place at the neutral stage. Ideally speaking while all collision induced processes occurring during neutralisation and reionisation should be constant in a series of experiments the unimolecular processes are expected to gain in intensity with increasing lifetime.In the next chapters some aspects of this approach will be outlined. 4 Variation of the neutrals’ lifetimes To a first approximation the same experimental set-up as used for NCR experiments [Fig. 6(b)] can be applied. Neutralisation in the first collision cell followed by deflection of all remaining ions with both deflector electrodes and reionisation of the neutral beam in the third cell (‘long distance’ LD-NR) affords a substantially longer flight path as compared to NR experiments using cells 2 and 3 together with the second deflector only [‘short distance’ SD-NR Fig.6(a)]. Comparison of these spectra allow us to identify the reactions occurring at the neutral stage. However some limitations should be mentioned. (i) The neutral beam cannot be guided by ion optics as ions can. Thus a longer flight path may result in larger mass discrimination effects than the shorter one. Direct comparison of SD- and LDNR mass spectra is therefore mostly restricted to ratios of fragments with similar masses. (ii) Often the time range is too short to observe fragmenting neutrals. For example the SD- and LD-NR mass spectra of the acepentalene system9 do hardly differ from each other indicating that the neutral does not decompose in the ms timescale. (iii) As the neutralisation step occurs in different collision cells it may be difficult to ensure that perfectly identical conditions prevail in both experiments.As a brief example let us return to methyl hydroperoxide23 for which Fig. 5 does not only show the NCR experiments but includes the SD-NR [Fig. 5(a)] and LD-NR (Fig. 5(b)] mass spectra. In good agreement with the NCR results the ratio of HCO+ and HO2 + fragments changes from the shorter (ca. 8 1) to the longer (ca. 11 1) flight path indicating cleavage of the 95 Chemical Society Reviews 1998 volume 27 Fig. 6 (a) Short distance (SD) NR experiment. Cell 1 is empty and a conventional NR experiment is performed using cell 2 and 3 together with the second deflector. (b) Long distance (LD) NR experiment. Neutralisation is achieved in cell 1 and reionisation in cell 3 while the second cell is empty and both deflectors are used.Fig. 7 Instrumentation used for variable-time neutralisation-reionisation experiments with a tandem quadrupole acceleration-deceleration mass spectrometer. After acceleration of the ions neutralisation occurs in a first collision cell and remaining ions are reflected by an electric potential. Subsequently reionisation occurs in a 60 cm conduit under single collision conditions. A system of electric potentials and a particular scan mode is used to distinguish reionised species formed in the desired region of the conduit from those generated downstream. (a) Short and (b) long flight paths. For details see ref. 24. peroxidic O–O bond as the predominant neutral reaction pathway.In 1994 Kuhns and Turecek24 introduced a new technique (Fig. 7) to vary the neutrals’ lifetimes in the NR experiments by using a tandem quadrupole acceleration-deceleration mass spectrometer equipped with an advanced collision cell device. After acceleration the ion beam is neutralised in the first collision cell and the unreacted ions are reflected by an electric potential. The second collision cell consists of a 60 cm flight tube the conduit segmented into four regions [Fig. 7(a) 1–4] and equipped with a system of electric potentials. The neutral beam is reionised throughout the whole flight tube. However by appropriately adjusting the electric potentials only the ions formed in one of the four segments of the conduit can be monitored. A short neutral flight path is realized in Fig.7(a). Ions formed in segment 1 floated at +250 V do not have kinetic energies which fit the voltage at the deceleration lens and at the Chemical Society Reviews 1998 volume 27 96 second quadrupole and are not registered. In contrast ions formed in segment 2 floated at a higher voltage which is scanned simultaneously with the deceleration lens and the second quadrupole are detected. By switching from one [Fig. 7(a)] of these regions to the next [Fig. 7(b)] the flight path of the neutrals is increased while that of the ions becomes shorter in the opposite sense. Only species reionised in segment 3 are monitored now. Comparison of the spectra obtained in this manner allows us to distinguish between neutral and ion fragmentations because those of the neutrals become more intense while ion decay decreases in intensity due to the shorter ion flight path.It is important to note that the collision gas is admitted uniformly to the flight tube at a pressure which guarantees single collision conditions (ca. 90% T). This ensures comparable reionisation conditions at each of the subunits of the conduit and avoids the problems connected with switching between different collision cells for neutralisation as described above for long and short distance NR experiments. Furthermore under these conditions ca. 95% of the reionised particles undergo only one collision which leads to reionisation. Thus unimolecular decompositions of the neutrals can be monitored while collision induced decompositions only contribute to a very minor extent.A kinetic analysis of ion and neutral fragmentations is possible using the variable-time NR technique provided that the relative reionisation cross sections are known or can be measured. This procedure yields phenomenological dissociation rate constants kn for neutral and ki for ionic species in the range of 104–106 s21.24 Much faster or slower fragmentations of the neutrals cannot be resolved within the 100 ns–5 ms timescale of the variable-time NR experiment. As an example let us refer to the measurement of the rate constant kn for the decay of a beam of methyl iodide CH3I which demonstrated that the resonant electron transfer using CH3I for neutralisation yields neutral methyl iodide with only modest excitation energy.In contrast off-resonance neutralisation of the CH3I+. beam with other target gases gave rise to higher rate constants which were rationalised by invoking initial formation of Rydberg states. Cl H Cl H S C C CH2 H3C Cl F 19• 17• 18• Some atmospherically important radicals i.e. CH3SCH2 . 17. CHClF. 18. and CHCl2 . 19. have been studied by this technique.25 For example the oxidation of dimethyl sulfide has been proposed to involve the reaction of HO. radicals with dimethyl sulfide and the methylthiomethyl radical 17. is believed to be an important intermediate. In line with results obtained from NR and NCR experiments variable-time NR experiments not only suggest 17. to be a stable species but also demonstrate that neutral 17.decomposes unimolecularly via methyl loss. As far as reaction kinetics are concerned the fragmentation of reionised 17. proceeds however at a higher rate constant as that of neutral 17 Similarly 18. and 19. have been reported to dissociate into CHF + Cl. and CHCl + Cl. respectively; both radicals may serve as sources for chlorine atoms relevant for ozone depletion in the upper atmosphere. Variable-time NR mass spectrometry has also been applied for investigations of fundamental issues in organic chemistry e.g. the protonation sites of pyridine.26 It is quite clear that the nitrogen atom represents the most basic position of pyridine followed by C(3) C(2) and C(4) with decreasing proton affinities.Combining isotopic labelling with neutralisationreionisation experiments allows us to distinguish between Nand C-protonation. The basic idea is depicted in Scheme 3. If the +e– D+ H H N N –D• D 20+ H H N N –D• +e– D+ D D D 20• N N H H 21+ –H• D N 21• Scheme 3 nitrogen atom in pyridine is deuteriated and cation 20+ subsequently neutralised exclusive loss of a deuterium atom from neutral 20. is expected. In contrast C-protonation may yield hydrogen and deuterium losses from intermediate 21 These experiments have of course to be cross-checked with protonated [2H5]pyridine in order to account for kinetic isotope effects. Experimentally exclusive N-protonation was found for weak Brønsted acids such as CH3NH3 + NH4 + tert-C4H9 + and H3O+ while a small fraction of C-protonated pyridine 21+ was formed in the reaction of pyridine with CH5 +.Another interesting feature of the NR mass spectra of protonated pyridines is the observation of recovery signals which demonstrate the existence of hypervalent pyridinium radicals to be stable species. Earlier the groups of Porter1 and species such as H3 . ND4 . and D3O. do exist in the gas phase. others27,28 provided evidence that a variety of hypervalent This field of NR mass spectrometry has recently been revisited by the variable-time NR and NCR techniques. For example several hypervalent ammonium radicals29 with different substitution patterns and tetracoordinated oxygenated phosphorus radicals30 have been investigated.For alkyl substituted 22. 23. N• P• N• P• HO CH3O OH OH H H OCH3 OCH3 22• 24• 25• 23• H H H N N N H H H 28• 27• 26• and 26.–28. homolytic cleavages of N–H N–CH3 or N–CH2Ph bonds prevail thus giving rise to the corresponding neutral amines. While the transient existence of these radicals has been proven by the observation of recovery signals for deuterated isotopologues of 22 and 23 no recovery signals were found in the +NR+ mass spectra of neither 26+–28+ nor their deuterated analogues. Similarly for the tetrahydroxy phosphonium radical 24 no a stable species for which the loss of H. (DHr = 27 recovery signal was detected.30 Theory however predicts 24 as 3O)3P and (CH3O)2P(OH) as neutral products.kcal mol21) and water (DHr = 22 kcal mol21) represent thermodynamically favourable fragmentation channels. These are indeed observed in the NR and NCR experiments carried out with P(OH)4 + cations. In contrast for the cationic P(OH)4 + precursor these two reaction pathways are endothermic by ca. 120 and 69 kcal mol21 respectively. Thus it is unlikely that the hydrogen atom and water losses occur at the cationic stage. Similarly 25 reacts via losses of H. and CH3 . yielding (CH for 25 also losses of HO. and CH3O. radicals are observed thus 3O)3PO and (CH3O)2(HO)PO as products. In contrast to 24 producing (CH Most recently variable-time NR experiments were perfor- 3S. radicals and its D and 34S labeled med with neutral H isotopologues.31 These experiments not only demonstrated that these exist as transient neutrals but also unraveled their reactivity.Hydrogen atom losses dominate which are affected by a normal kinetic isotope effect. As observed earlier1,7 for other hypervalent radicals neutralisation of vibrationally excited precursor ions leads to an increase of the recovery signal intensity. This indicates that hot precursor ions due to more favourable Franck–Condon factors have a higher probability of being neutralised into stable states of the hypervalent radicals. 5 Photoexcitation of transient neutrals The metastability of hypervalent radicals due to the presence of bound excited states has been under intense discussion not the least because of some inconsistencies in the interpretation of the data.While theory in some cases predicted unstable or very weakly bound ground states the experiments showed hypervalent radicals to exist.1 In order to probe further this aspect irradiation of the neutral species with photons of wavelengths suitable for the ionisation of metastable excited states but not the ground states seems a promising strategy and an appropriate experiment was designed by the group of Turecek.32 97 Chemical Society Reviews 1998 volume 27 Fig. 8 Instrumentation for laser photoionisation experiments an Ar laser (488 and 514.5 nm) is directed antiparallel to the flight axis of the neutral particle beam. The neutrals are allowed to interact with the photons throughout the whole length of the conduit (60 cm).They equipped their tandem quadrupole acceleration-deceleration mass spectrometer shown in Fig. 7 with a laser that is directed antiparallel to the axis of the particle beam (Fig. 8). 4 . radicals with photons in 4 + Upon photolysis of metastable ND the range of 2.41–2.54 eV only a very minor fraction of ND is formed.32 As the ground state bears an ionisation energy of 4.6 eV these ions cannot be generated from ground state ND4 Rather a small amount of excited ND4 . was generated in the neutralisation step. In contrast the signals corresponding to ND 4 . and subsequent reionisation of the 3 and ND2 . are much more intense owing to photodissociation of ground state ND 3 and ND2 . fragments. This rationale is in agreement with ND the potential energy diagram depicted in Fig.9 for NH4 The Fig. 9 Potential-energy diagram (kcal mol21) for NH4 2A1 ground state is prevented by a barrier of ca. 0.41 eV (ca. 9.5 kcal mol21) from dissociation into NH3 + H The barrier for the loss of H2 with concomitant formation of NH2 . in its 2A1 state is even higher but still below the excitation limit given by the photon energy. In contrast to the ground state the excited 2T2 state is kinetically much more stabilised. However dipoleallowed transitions to lower states are possible and the 2T2 ? 2A1 transition may explain that only a minor fraction of excited 4 . survived the time delay between neutralisation and ND photoionisation. Chemical Society Reviews 1998 volume 27 98 3)2O + D.or CH3OD + oxonium radicals (CH3)2OD. 2932 This species is found A distinctly different picture is found for dimethyldeuterioexperimentally to be metastable against the thermodynamically favourable dissociations into either (CH CH3 . and metastable 29. is easily photoionised by 2.41–2.54 eV photons. Theory predicts the 2AA ground state of 29. to be repulsive with respect to dissociation of the O–H(D) bond (Fig. 10). However the excited 2AB electromer which is characterised as a 3p Rydberg state is expected to be metastable. Its vertical ionisation energy amounts to 2.22 eV which matches the range of the photon energies. Fig. 10 Potential-energy diagram (kcal mol21) for ground and excited states of (CH3)2OH. 29 These few examples may suffice to demonstrate the performance of a combination of NR experiments with photochemistry for the examination of metastable excited states.It should be noted however that one has to account for multi-photon ionization when using this technique. For the detection of high Rydberg states which are transparent for the Ar laser beam field ionisation provides an attractive alternative.3 Thus the possibility to perform photoionisation or photodissociation experiments of transient neutrals helps to bridge the gap between ground and Rydberg state chemistry. Application of this method to larger systems may be of quite some interest for future work. 6 Neutral and ion decomposition difference (NIDD) mass spectrometry In the previous sections we have described how the unimolecular decay of neutrals generated by neutralisation-reionisation mass spectrometric experiments can be probed by several techniques.However these methods require rather special experimental set-ups which are not common to most of the laboratories concerned with the generation and investigation of elusive neutrals. In this section a new approach coined ‘Neutral and Ion Decomposition Difference’ mass spectrometry (NIDDMS) 33,34 will be described which can be performed with a conventional NR set-up. Its scope will be demonstrated by applying it to systems as different as the formate anion peroxide cation radicals the .CH2COO2 distonic anion and alkoxy radicals. Fig. 11 (a) Conventional NR experiment with charge inversion of the ion beam (2NR+ or +NR– respectively).(b) Charge reversal experiment (2CR+ or +CR2). Recently the femtosecond dynamics for the formation of three membered Ag3 rings from its linear form have been studied with a related method.35 The Negative ion–Neutral atom–Positive Ion (NeNePo) charge reversal spectroscopy relies on the photon-induced vertical neutralisation of an anion e.g. linear Ag3 2. This structure represents a transition structure on the neutral potential-energy surface. The ring closure reaction can be monitored by a time-delayed second laser pulse for cationisation of the neutral. In contrast the NIDD approach uses collisions instead of laser pulses for neutralisation and reionisation. Here variation of the lifetimes of the neutral species is achieved by comparing conventional 2NR+ or +NR2 mass spectra with the corresponding charge reversal36 mass spectra (2CR+ or +CR2 respectively see Fig.11). While the NR procedure necessarily requires two separate collision events and intermediate deflection of the unreacted ions charge inversion in CR experiments is afforded in a single collision by a two-electron transfer. Ideally speaking the CR signals are due to ionic fragmentations whereas in the NR spectra dissociations of ionic and neutral processes are superimposed. Normalisation33,34 of the spectra to the sum of all fragments followed by subtraction of the CR from the NR intensities [eqn. (3)] results in NIDD spectra in which processes due to the fragmentations of neutrals have positive intensities while ionic contributions show up as negative peaks thus providing firm criteria for the identification of neutral and ionic processes.Ii(NIDD) = [Ii(NR)/SiIi(NR)]2[I3(CR)/SiIi(CR)] (3) The NIDD method has the major advantage that it uses a standard NR set-up. Thus both experiments can be conducted rather quickly one following the other so that ion source conditions and ion focus parameters can be kept constant during the measurements. As will be shown below NIDD is applicable to the reactivity of open- and closed-shell neutrals and also to both directions of charge inversion. 2]. system in As a first example let us discuss the [C,H,O some detail including the results of a theoretical study of the respective potential-energy surfaces.33,34 The two ionic precursors used are typical organic ions namely the formate ion HCOO2 302 and the hydroxyacylium ion HOCO+ 31+ viz.protonated carbon dioxide. The 2NR+ and 2CR+ spectra of formate ion 302 [Fig. 12(a,b)] are fairly different from each other. The signals for the recovery ion and the HCO+ fragment are barely visible in the 2NR+ spectrum but quite intense in the 2CR+ experiment. Due to these pronounced differences in the 2NIDD+ spectrum [Fig. 12(c)] CO2 +. is the dominant feature on the positive scale. Thus we conclude that the formoxyl radical HCOO. formed upon electron detachment from HCOO+ dissociates to a large amount by loosing a hydrogen atom i.e. HCOO ? CO2 + H. while the HCO+ fragment arises from the cationic surface.Interestingly in the spectra of DCOO2 30a2 [Fig. 12(d,e)] the recovery signals and also the DCO+ fragments are much more intense thus pointing to a sizeable kinetic isotope effect on the stability of the transient formoxyl radical formed upon neutralisation. In contrast the loss of an oxygen atom to yield HCO+ or DCO+ respectively is not subject to isotope effects. The experimental results are fully confirmed by the potential-energy surfaces of [C,H,O2]+/./2 systems calculated at the Becke3LYP/ 6-311++G(d,p) level of theory (Fig. 13). The barrier for the hydrogen atom loss from HCO2 . (TS 30.) is expected to be small. Reionisation of the neutral as well as charge inversion of the anion does most likely not involve the singlet HCO2 + cation but rather the triplet analogue which is calculated to be formed in the 2CR+ experiment with only a negligible amount of excitation energy due to similar geometries.This accounts for the more abundant survivor ion current observed in the 2CR+ experiment. From triplet HCO2 + the O-atom loss corresponds to the most favourable exit channel in line with the experiment. Unfortunately the +NR2 and +CR2 mass spectra of protonated carbon dioxide HOCO+ are quite similar and therefore cannot be analysed with the NIDD scheme because in this case the absolute signals in the +NIDD2 spectrum vanish within experimental error. As mentioned above the ionised methyl hydroperoxide bears an intact peroxide connectivity,23 and the same holds true for the neutral species (see above).Indeed the +NIDD2 mass spectrum of CH3OOH+. (Fig. 14) shows positive signals for the HO2 and CH3O2 fragments due to O–O bond cleavage in neutral methyl hydroperoxide. In contrast all fragments due to C–O or H–O bond ruptures are negative indicating their ionic origin. Overall the reactivity of peroxides33,37 generated by collisional neutralisation of the cations follows expectation and demonstrates the applicability of +NIDD2 for the examination of the reactivities of transient closed-shell neutrals in the gas phase. Similarly the +NIDD2 mass spectrum of ionised hydrogen peroxide [Fig. 15(c)] shows an intense positive peak for the formation of HO. radicals again indicating O–O bond cleavage in agreement with the conventional HOOH connectivity.However the +NR2 and +CR2 mass spectra [Fig. 15(a,b)] of [H2,O2]+. show a distinct recovery signal which was not observed for the methyl substituted homologues. By detailed experimental and theoretical investigation including [H,D,O2] and [D2,O2] the recovery signal was shown to be due to the long sought after singlet water oxide molecule H2OO as transient 99 Chemical Society Reviews 1998 volume 27 Fig. 12 (a) 2NR+ mass spectrum (O2 80% T; O2 80% T) (b) 2CR+ mass spectrum (O2 80% T) and (c) 2NIDD+ spectrum of formate ion HCOO2 302 generated by chemical ionisation of formic acid with N2O as reagent gas. (d)–(f) Analogous spectra for the deuteriated ion DCOO2 30a2. neutral.38 Thus in line with other experiments such as NR/CA NCR SD-/LD-NR it was demonstrated that the neutral beam contains a small amount of the H2OO isomer within an excess of HOOH molecules.Nevertheless neutral water oxide could be identified by its ability to form stable anions. The calculated potential-energy surface of neutral [H2,O2] (Fig. 16) reveals that the bulk generation of water oxide is not straightforward at all and only sophisticated techniques like NR mass spectrometry can be used to probe its existence at a molecular level. Thus this example demonstrates once more the enormous potential of the NR technique for the generation of highly reactive and elusive neutrals which are difficult to produce and detect by any other spectroscopic means. Mass spectrometric means are particularly favourable for the identification of neutral water oxide since it relies on the HOOH+.? H2OO+. isomerisation of a small amount of the ionised hydrogen peroxide precursors from which vertical ionisation into the two neutral minima is possible. One of the most interesting classes of compounds in gasphase chemistry are the distonic ions i.e. open-shell ions in which radical and charge centres are spatially separated. While these ions can exhibit rather unusual geometric structures they are in fact quite often thermochemically or kinetically more stable than the conventional ionic counterparts e.g. the .CH2OH2 +/CH3OH+. couple. Distonic ions attract on-going attention and the same should apply for the corresponding neutrals because unless rearrangements occur these should be formed as ylides from a-distonic ions or as biradicals from b- or higher distonic ions.As a representative for a negatively charged b-distonic ion let us discuss the acetoxy anion 2COO2 32+39 The 2NIDD+ spectrum (Fig. 17) shows .CH significant positive signals for the CO2 +. CH2O+. and CO+. fragments which can be interpreted in terms of Scheme 6. Electron detachment from .CH2COO2 leads to the corresponding diradical 32 either in singlet or triplet states. The neutral triplet then dissociates in the ms timescale into 3CH2 and CO2 which are subsequently reionised. The singlet however can cyclise to a-aceto lactone 33 which gives rise to the CH2O+. and CO+. fragments i.e. either the excess energy gained in recombination leads to dissociation at the neutral stage followed by reionisation of the neutrals or these fragments are formed upon reionisation of 33 but not 32.Recent results from our laboratory show that oxyallyl anion radicals .CH2C(O)CH2 2 Chemical Society Reviews 1998 volume 27 100 give rise to positive CO and C2H4 signals in the 2NIDD+ spectrum. This indicates that ring closure occurs analogously at the neutral stage. Thus 2NIDD+ of radical anions is particularly suitable to examine biradicals which are rather difficult to produce by other means. One particular feature may need further comment according to Scheme 4 the CH2 +. fragment is O –e– singlet H2C O O O 33 NR –e– O• O– H2C H2C –e– 32 3CH2 + CO2 triplet O CH2 +• + CO2 O H2C 32+ Scheme 4 CH2O+•/CO+• CH2 +•/CO+• 2 +.32– CR –2e– due to fragmentation of neutral 32 and should thus appear on the positive part of the 2NIDD+ spectrum yet it gives rise to a negative signal. According to the NIDD scheme CH therefore originates from ionic species and a facile rationalisation is the direct C–C bond cleavage of the transient 2COO+ cation formed upon charge inversion in which the .CH methylene fragment is more likely to be charged because the ionisation energies of CH differ. Hence the negative signal for CH2 +. is not in 2 (10.4 eV) and CO2 (13.8 eV) largely contradiction with the NIDD scheme because more CH2 +. is formed by direct dissociation of the cations in the 2CR+ experiment than by reionisation of the methylene fragment in the 2NR+ process.This and similar compensation effects must however be accounted for in the interpretation of NIDD spectra. As a final example the gas-phase reactions of neutral alkoxy radicals generated by neutralisation of alkoxide ions will be discussed.40 Alkoxy radicals have been chosen because in condensed matter their reactivity is well understood and they may serve as suitable test systems for the performance of NIDD. Alkoxy radicals with short and/or branched hydrocarbon chains e.g. methoxy ethoxy isopropoxy or tert-butoxy radicals, Becke3LYP/6-311++G(d,p) level of theory. undergo a-cleavages [eqns. (4–9)] which give rise to the losses of H. and CH3 . respectively.The intensity ratios observed for C–H and C–C bond cleavage products for ethoxy and isopropoxy radicals are consistent with thermochemistry with the least endothermic process being favoured in the NIDD spectra.40 CH3O. Fig. 15 (a) +NR2 mass spectrum (Xe 80% T; benzene 70% T) (b) +CR2 mass spectrum (benzene 70% T) and (c) +NIDD2 spectrum of hydrogen peroxide cation radicals generated by electron ionisation of H2O2. ?CH2O + H. CH3CH2O. ?CH3CHO + H. Fig. 16 Potential-energy surfaces of neutral singlet and triplet [H2,O2]. 3 . Heats of formation are given in kcal mol21. (4) (5) (6) (7) 3CH2O. ?CH2O CH + CH (CH3)2CHO. ?(CH3)2O + H. Fig. 13 Potential-energy surfaces of [C,H,O2]+/./2 species calculated at the Fig. 14 +NIDD2 mass spectrum of methyl hydroperoxide generated by electron ionisation.Fig. 17 2NIDD+ mass spectrum of .CH2COO2 distonic anion 322 101 Chemical Society Reviews 1998 volume 27 Fig. 18 (a) 2NR+ mass spectrum (O2 80% T; O2 80% T) (b) 2CR+ mass spectrum (O2 80% T) and (c) 2NIDD+ spectrum of pentanolate ions 342 generated by negative ion chemical ionisation of n-pentan-1-ol with N2O. (8) (CH3)2CHO. ?CH3CHO + CH3 . (CH3)3CO. ?(CH3)2O + CH3 . (9) Alkoxy radicals with longer side-chains exhibit remarkably different reactivities. Thus 1,5-hydrogen migrations become feasible and indeed Barton-type rearrangements are observed. The pentyloxy radical 34. may serve as an example.40 While the 2NR+ and 2CR+ spectra of 342 [Fig. 18(a,b)] are rather complicated and difficult to analyse the 2NIDD+ mass spectrum of pentoxide ions 342 [Fig.18(c)] reveals only three distinct signals on the positive scale corresponding to CH2OH+ its C4H8 +. counterpart and the loss of water indicating a hydrogen migration from the alkyl chain to the oxygen centered radical. The 2NIDD+ mass spectra of deuterium labeled pentoxide ions revealed the occurrence of a clean 1,5-H shift,40 suggesting the mechanism proposed in Scheme 5 which O– C5H9 + –H2O –e– O H O• OH 35• C4H8 + •CH2OH –e– –e– C CH2OH+ 4H8 +• Scheme 5 34– –e– 34• involves a six-membered transition structure for hydrogen migration with 35. being the central intermediate from which Chemical Society Reviews 1998 volume 27 102 upon reionisation two reaction pathways lead to either the formation of CH2OH+ and C4H8 +.or to the water loss. Appropriate substitution of the hydrocarbon backbone even permits us to examine the stereochemical features of the 1,5-hydrogen migration in transient alkoxy radicals. To this end a methyl group at the C(3) position of a pentyloxy radical was introduced as a stereochemical marker. The stereoselectivity of the hydrogen migration can be monitored using the diastereospecifically labeled [4-2H]-3-methylpentyloxy radicals 36a. and 36b. (Scheme 6). The 2NIDD+ spectra of the H D O– – e– H3C H 36a– • D O H H ~H D OH • 37a• CH2OH –e– CH2OH+ diastereoisomers show distinctly different ratios of CH2OH+ versus CH2OD+ formation i.e.6.4 1 for the isotopologue 36a2 as compared to 0.5 1 for the diastereoisomer 36b2. Considering that the only difference between 36a2 and 36b2 is the relative stereochemistry at the C(3) position the striking diastereoselectivity has to be traced back to a stereochemical differentiation of the transition structures in the course of the 1,5-hydrogen (deuterium) migration. Of course in addition to a stereochemical effect (SE) the measured CH2OH+–CH2OD+ ratios for 36a2 and 36b2 are also affected by the operation of a kinetic isotope effect (KIE). Assuming a combined action of a steric and a kinetic effect it is possible to semi-quantitatively estimate the magnitude of these two effects by applying a simple algebraic approach.33,40 As a result from the measured intensities of CH2OH+ and CH2OD+ one can derive the two relevant components with KIE = 1.8 ± 0.3 and SE = 3.6 ± 1.3 for the 1,5-hydrogen atom transfer in neutral 3-methylpentyloxy radicals.A straightforward analysis of these results can be accomplished by the mechanism depicted in Scheme 6 for the diastereoisomer 36a2. Upon electron detachment from the anion 36a2 the neutral radical 36a2. is formed which can subsequently adopt conformations with suitable geometries for the 1,5-hydrogen (deuterium) transfer involving the C(4) position. Thus two different chair-like transition structures are possible leading to H- and D-atom transfer respectively. As a result the carbon centered radicals 37a.and 37aA. are formed which subsequently decompose and after reionization of the neutral hydroxymethyl fragments yield the corresponding 37a¢• CH2OD –e– CH2OD+ H D H D O– O• H CH3 H3C H 36b– • O D H H ~D H OD • 36a• Scheme 6 CH2OH+ and CH2OD+ cations. For the diastereoisomer 36a. the geometry of the TS associated with migration of a hydrogen atom is a chair-like structure in which both methyl substituents adopt favourable equatorial positions. In contrast deuterium transfer from 36a. has to proceed via a transition structure in which one of the methyl groups is forced into the energetically less favourable axial position. However a theoretical investigation of the n-pentyloxy parent system revealed that this model is idealized in that it assumes a perfect chair-like transition structure for the 1,5-hydrogen migration neglecting the partial linearization of the C(4)–H–O moiety (see above).The Newman-type representations as depicted in Scheme 7 clarify Scheme 7 the influence of this effect. For example in the transition structure corresponding to the migration of a hydrogen atom (Scheme 6) the 1,2-interactions of the two methyl groups at C(3) and C(4) are reduced upon linearization of the C(4)–H–O subunit. In contrast for the deuterium shift these steric interactions are augmented as compared to an ideal chair-like structure. Both pictures the conformational analysis of chairlike transition structures as well as the description in terms of 1,2-interactions of the methyl groups predict that the isotopologue 36a.will preferentially undergo hydrogen-atom transfer because this path is favoured by the reduced steric interaction as well as by the kinetic isotope effect thus giving diastereoisomer 36b. with an inverted relative stereochemistry rise to a large CH2OH+–CH2OD+ ratio. In contrast in the at the C(3) position deuterium-atom transfer will be conformationally favoured but kinetically hampered; consequently SE and KIE almost level out for this stereoisomer as is indeed observed experimentally. 7 Conclusions The present review article focusses on techniques for the investigation of the chemical reactivity of neutrals generated in NR experiments. As an underlying guideline a roughly chronological order of the development of these methods has been followed.Parallel to the methodological progress the chemical systems under study became more and more complex. Beginning with a qualitative examination of simple bond cleavages such as CH3CO. ? CH3 . + CO and CH3Cl ? CH3 . + Cl. NR mass spectrometry now allows for the quantitative evaluation of kinetic data kinetic isotope effects and even the regio- and stereo-selectivity of a chemical reaction. It is no longer a method for merely probing the existence and structure of elusive neutrals but has developed to a useful tool for mechanistic studies on the gas-phase reactivity of neutrals in high vacuum. However the examination of neutral species with NR techniques is always indirect and remains hampered by a few limitations.(i) As mass analysis of the neutral beam is not possible the resulting spectrum always consists of a superposition of reactions of both charged and neutral species. (ii) Every NR experiment necessarily involves collisions with target gases. Thus terms like ‘unimolecular’ and ‘collision induced’ become somewhat vague. (iii) In NR experiments the internal energies of the neutrals do not follow a Maxwell– Boltzmann distribution which is mentally anchored in and guides intuition of most chemists. Rather Franck–Condon factors determine the population of electronic and vibrational states. Nevertheless for the time being the NR methodologies described here in its numerous variants represent the best approach for the generation and examination of many elusive neutrals and their reactivities.If there remain uncertainties as far as the reactivity of a neutral is concerned due to the particular nature of the experiment a combination of several approaches as discussed above is strongly recommended. Most of the work presented in this review article took advantage from the synergy of theory and experiment although this has not been followed in detail here due to the lack of space. It should be stressed that theoretical treatment of the isolated neutrals under study of the transition structures connecting them their energies and their geometries provides a great amount of additional information. Furthermore questions concerning electronic states electron affinities and ionisation energies can be addressed and reaction pathways can be modeled.As the calculations refer to isolated molecules a direct link to the experiments that are performed in the highdiluted gas phase is established. Thus theory provides an extremely helpful complementary tool for the investigation of elusive gaseous neutrals. In conclusion as repeatedly demonstrated NR mass spectrometry coupled with computational work has an outstanding potential for the generation and characterisation of unconventional neutrals. With the more recent development of a variety of methods for the investigation of their chemical reactivity a new field is opened up in this branch of chemistry. The authors of this Review believe that the progress in this field will not only lead to new insight into chemical questions concerning the properties of elusive neutrals but also to the design of new experimental techniques beyond those described in this contribution.8 Acknowledgements We are grateful to Professor C. Wesdemiotis and Professor W. C. Lineberger for communicating some unpublished results and to Professor J. K. Terlouw for helpful discussions. We thank the Deutsche Forschungsgemeinschaft the Gesellschaft der Freunde der Technischen Universit�at Berlin and the Fonds der Chemischen Industrie for financial support. 9 References 1 G. I. Gellene and R. F. Porter Acc. Chem. Res. 1983 16 200. 2 C. Wesdemiotis and F. W. McLafferty Chem. Rev. 1987 87 485. 3 J. L. Holmes Mass Spectrom.Rev. 1989 8 513. 4 A. W. McMahon S. K. Chowdhury and A. G. Harrison Org. Mass Spectrom. 1989 24 620. 5 F. W. McLafferty Science 1990 247 925. 6 F. Turecek Org. Mass Spectrom. 1992 27 1087. 7 N. Goldberg and H. Schwarz Acc. Chem. Res. 1994 27 347. 8 D. V. Zagorevskii and J. L. Holmes Mass Spectrom. Rev. 1994 13 133. 9 R. Haag D. Schr�oder T. Zywietz H. Jiao H. Schwarz P. v. R. Schleyer and A. de Meijere Angew. Chem. 1996 108 1413; Angew. Chem. Int. Ed. Engl. 1996 35 1317. 10 G. A. McGibbon J. Hru�s �ak D. J. Lavorato H. Schwarz and J. K. Terlouw Chem. Eur. J. 1997 3 232 and references cited 11 D. Lahem R. Flammang and M. T. Nguyen Chem. Phys. Lett. 1997 12 M. J. Polce Y. Kim and C. Wesdemiotis Int. J. Mass Spectrom. Ion 13 M.W. Wong C. Wentrup E. H. Mørkved and R. Flammang J. Phys. 14 V. Q. Nguyen and F. Turecek J. Mass Spectrom. 1996 31 843 and therein. 270 93 and references cited therein. Processes in press. Chem. 1996 100 10536 and references cited therein. references cited therein. 103 Chemical Society Reviews 1998 volume 27 15 R. Feng C. Wesdemiotis M. A. Baldwin and F. W. McLafferty Int. J. Mass Spectrom. Ion Processes 1988 86 95. 16 Review J. K. Terlouw Adv. Mass Spectrom. 1989 11 984. 17 M. J. Polce �S. Beranov�a M. J. Nold and C. Wesdemiotis J. Mass Spectrom. 1996 31 1073. 18 S. K�ornig J. H. M. Beijersbergen and J. Los J. Phys. Chem. 1990 94 611 and references cited therein. 19 C. E. C. A. Hop and J. L. Holmes Int. J. Mass Spectrom. Ion Processes 1991 104 213.20 C. Wesdemiotis R. Feng P. O. Danis E. R. Williams and F. W. McLafferty J. Am. Chem. Soc. 1986 108 5847. 21 F. Turecek D. E. Drinkwater A. Maquestiau and F. W. McLafferty Org. Mass Spectrom. 1989 24 669. 22 C. Wesdemiotis B. Leyh A. Fura and F. W. McLafferty J. Am. Chem. Soc. 1990 112 8655 and references cited therein. 23 C. A. Schalley D. Schr�oder and H. Schwarz Int. J. Mass Spectrom. Ion Processes 1996 153 173. 24 D. W. Kuhns and F. Turecek Org. Mass Spectrom. 1994 29 463. 25 M. Sad�ýlek and F. Turecek J. Phys. Chem. 1996 100 224 and references cited therein. 26 V. Q. Nguyen and F. Turecek J. Mass Spectrom. 1997 32 55 and references cited therein. 27 M. Sirois M. George and J. L. Holmes Org. Mass Spectrom. 1994 29 11. 28 S. A. Shaffer and F. Turecek J. Am. Soc. Mass Spectrom. 1995 6 1004. 29 S. A. Shaffer and F. Turecek J. Am. Chem. Soc. 1994 116 8647. Chemical Society Reviews 1998 volume 27 104 30 F. Turecek M. Gu and C. E. C. A. Hop J. Phys. Chem. 1995 99 2278. 31 M. Sad�ýlek and F. Turecek J. Phys. Chem. 1996 100 15027. 32 M. Sad�ýlek and F. Turecek J. Phys. Chem. 1996 100 9610; Chem. Phys. Lett. 1996 263 203. 33 C. A. Schalley Gas-Phase Ion Chemistry of Peroxides Ph.D. Thesis (Technische Universit�at Berlin 1997 D 83) Shaker Herzogenrath/ Germany 1997. 34 C. A. Schalley G. Hornung D. Schr�oder and H. Schwarz Int. J. Mass Spectrom. Ion Processes in press. 35 D. W. Boo Y. Ozaki L. H. Andersen and W. C. Lineberger J. Phys. Chem. A 1997 101 6688 and references cited therein. 36 M. M. Bursey Mass Spectrom. Rev. 1990 9 555. 37 C. A. Schalley A. Fiedler G. Hornung R. Wesendrup D. Schr�oder and H. Schwarz Chem. Eur. J. 1997 3 626. 38 D. Schr�oder C. A. Schalley N. Goldberg J. Hru�sak and H. Schwarz Chem. Eur. J. 1996 2 1235. 39 D. Schr�oder N. Goldberg W. Zummack H. Schwarz J. C. Poutsma and R. R. Squires Int. J. Mass Spectrom. Ion Processes 1997 165/166 71. 40 G. Hornung C. A. Schalley M. Dieterle D. Schr�oder and H. Schwarz Chem. Eur. J. 1997 3 1866. Paper 7/05838A Received 11th Augus

ISSN:0306-0012    

DOI:10.1039/a827091z

出版商:RSC    

年代:1998

数据来源: RSC

摘要:

Dibenzotetraaza[14]annulenes versatile ligands for transition and main group metal chemistry Philip Mountford† Department of Chemistry The University of Nottingham Nottingham UK NG7 2RD This article presents an overview of the chemistry of metal complexes of dibenzotetraaza[14]annulene ligands and highlights most of the recent developments. The title tetraazamacrocycles are related to the porphyrins but have a smaller N4 coordination cavity ‘hole size’ and typically possess a non-planar saddle-shaped conformation. The similarity of the dibenzotetraaza[14]annulenes to the porphyrins means that these synthetic macrocycles are of bioinorganic relevance while their distinctive individual characteristics make them interesting ligands in their own right.In early transition metal chemistry for example the dibenzotetraaza[14]annulenes have been studied as potential alternatives to the ubiquitous bis(h-cyclopentadienyl) ligand set while later transition metal derivatives can mimic certain biological systems and act as precursors to electroactive polymeric films. The dibenzotetraaza[14]annulenes have also recently allowed advances in structural and reactivity studies of main group organometallic and coordination compounds including the stabilisation of new metal–chalcogenide multiple bonds. 1 Introduction There is a tremendous interest in exploring new ligand environments for transition and main group metal chemistry and for developing synthetic mimics of biological systems. One class of ligand that has stimulated continuing interest in a wide range of areas for nearly thirty years are the dibenzotetraaza[ 14]annulenes [Fig.1(a)]. The chemistry of macrocyclic ligand complexes in general represents a vast area of research ranging over many areas of chemistry and biochemistry and the reader is referred to Lindoy’s excellent book for a succinct introduction to this topic.1 In the context of this current review Philip Mountford (b. 1962) gained a BSc (Hons) degree at Hatfield Polytechnic (1982–1986) and a DPhil degree at the University of Oxford (1986–1989) under the supervision of Professor M. L. H. Green. He remained at Oxford as a Junior Research Fellow at Wolfson College (1989–1992) Departmental Demonstrator at the Inorganic Chemistry Laboratory (1990–1992) and College Lecturer at Keble College (1990–1992).In 1992 he was appointed to a lectureship at the University of Nottingham. His research interests centre around studies of the synthesis structure bonding and reactivity of organometallic and coordination compounds and he is author or coauthor of over 80 publications in this area. † E-mail philip.mountford@nottingham.ac.uk; http://www.nottingham. ac.uk/ ~ pczwww/Inorganic/PMount.html it suffices to point out that macrocyclic ligands are considerably attractive in the quest for new chemistry not least because they can offer a wide variety of donor atom types ligand charges coordination numbers and resultant complex geometries. Macrocyclic complexes can also exhibit enhanced kinetic and thermodynamic stability in comparison with those of unidentate ligands.The purpose of this article is to provide the reader with a general introduction to the synthesis and properties of transition and main group metal compounds of dibenzotetraaza[ 14]annulene ligands and to highlight the more recent developments in their chemistry. It is not within the scope of Chem. Soc. Rev. to provide a full and comprehensive account of this area but it is hoped that this article will provide a suitable springboard for further reading. A detailed review of the transition metal chemistry of the Me4taa ligand [Fig. 1(a)] up to 1990 has been published.2 2 Typical features of dibenzotetraaza[14]annulenes and their complexes 2.1 Opening remarks We shall begin by examining the characteristic important features of the ligands and their metal complexes.Fig. 1 illustrates the general structure of dibenzotetraaza[14]annulenes together with that of the naturally-occurring macrocycle porphyrin with which they are often compared. Abbreviations used herein for some of the more common dibenzotetraaza[14]annulenes are also provided in Fig. 1 for ease of reference. Note that in some of the literature the alternative abbreviations used for the Me4taa and Me8taa ligands are ‘tmtaa’ and ‘omtaa’ respectively. It is instructive to compare and contrast the two types of tetraazamacrocycle. The dibenzotetraaza[14]annulene and porphyrin ligands both have R1 Abbrev. R1 R2 R3 R2 R2 H H H H2taa N N R3 R3 H H2Me4taa H Me H N H R3 N R3 R2 R2 R1 ( a) N H N Porphyrin N H N H2Me8taa H Me Me 105 ( b) Fig.1 Comparison of the dibenzotetraaza[14]annulenes (showing the abbreviations used in this review) and porphyrin Chemical Society Reviews 1998 volume 27 four coplanar nitrogen donor atoms can readily be deprotonated to form a dianion and possess a number of double bonds in the ligand framework. However the dibenzotetraaza[14]annulenes are H�uckel anti-aromatic (4n) whereas the porphyrins are fully delocalised aromatic (4n + 2) systems. Furthermore structural and theoretical (see Section 2.3) studies of dibenzotetraaza[ 14]annulenes and their complexes have shown that there is negligible delocalisation between the two o-phenylene and the 1,3-diiminato fragments.This is apparently due to the aromatic stability of the individual benzenoid rings and so p-delocalisation is confined within (but not between) the two types of fragment. Consequently in the deprotonated dibenzotetraaza[ 14]annulene dianions the negative charges tend to be localised on the 1,3-diiminato linkages [Fig. 2(a)] whereas in H H Me Me Me Me – N N N N H N N N N – Me Me Me Me ( b) ( a) H H Me Me Me Me – N N N N H N N N N Me Me Me Me H H ( d) ( c) Fig. 2 (a) Dianionic (b c) neutral and (d) monoanionic forms of tetramethyl dibenzotetraaza[14]annulenes. See the text for further details. porphyrins the charge can be more delocalised over the entire ligand.The flexibility (i.e. ability to adopt different conformations) of the dibenzotetraaza[14]annulene ligands may be placed somewhere between the essentially rigid delocalised porphyrins and the fully-saturated [16]aneN4 cyclam systems. 2.2 Ligand conformations and metal coordination geometries Further important differences between the dibenzotetraaza[ 14]annulene and porphyrin systems concern the ‘hole size’ of the macrocycle coordination cavity (defined as the average distance from N to the mid-point of the four N atoms) and the geometry of the ligand framework. As indicated by their name the dibenzotetraaza[14]annulenes have only a fourteen-membered inner ring (as compared to the larger sixteen-atom porphyrin inner ring).This gives rise to an N4 coordination cavity hole size of between 1.90 and 1.93 Å for H2taa and H2Mentaa which is about 0.1 Å less than that of the porphyrins. This feature is one of the main reasons (but also see below) why many dibenzotetraaza[14]annulene complexes have the metal lying out of the N4 plane whereas their porphyrin analogues tend to have the metal coplanar with the N4 donors. The second special characteristic of the dibenzotetraaza[ 14]annulene ligands is their tendency to adopt a so-called ‘saddle-shaped’ conformation when coordinated to a metal centre. This feature is especially pronounced for Me4taa and other homologues where R2 = Me because of steric interactions between the R2 methyl substituents and the hydrogen atoms of the o-phenylene rings.For Mentaa (n = 4 or 8) the non-planarity even persists in the free ligands as illustrated by Fig. 3(a) and (b) which show the structures of H2taa (planar) and H2Me4taa Chemical Society Reviews 1998 volume 27 106 Fig. 3 The planar and saddle-shaped solid state structures of (a) H2taa3 and (b) H2Me4taa.2 Hydrogen atoms are omitted for clarity and the remaining atoms are drawn as spheres of arbitary radius. respeively.2,3 The steric constraints imposed by the iminato methyl substituents are very important because they also contribute to the tendency for complexes of Me4taa and its homologues to have the metal atom lying out of the N4 plane (as mentioned above). This in turn leads the metal centres to favour five-coordination over six-coordination.Furthermore the non-planarity of Me4taa and other systems where R2 ­ H is believed to inhibit the types of intermolecular interactions in the solid state that give rise to the typically poor solubility of unsubstituted porphyrin and phthalocyanine complexes. Two other important factors affect the degree of nonplanarity of the macrocyclic ligands in their metal complexes. These are the radius and d-electron count (for transition elements) of the metal ion itself and to a lesser extent the number and type of any axial ligands. Hence even complexes of the parent taa ligand [Fig. 1(a)] which itself is planar in the solid state as H2taa can adopt a slightly saddle-shaped conformation depending on the identity of the metal and any ancilliary ligands present.For a given type of dibenzotetraaza[14]annulene complex the extent to which a metal atom is displaced from the N4 plane depends primarily on the metal radius and ligand field stabilisation effects. For example in the MIII complexes [M(Me4taa)X] (M = Co X = I; M = Fe X = Cl) the Co and Fe atoms lie 0.234 and 0.600 Å from the N4 plane respectively reflecting the larger effective radius of high-spin (S = 5/2) FeIII.2 The structure of the iron complex is shown in Fig. 4(a) by way of example and demonstrates clearly the displacement of the Fe atom from the N4 plane. Note that in [Co(Me4taa)I] the metal atom is apparently small enough to lie in the N4 plane and therefore it may be the inherent non-planarity of Me4taa and concomitant redirection of the N atom lone pairs out of the plane that causes the 0.234 Å displacement in this case.The remaining pair of structures in Fig. 4 illustrate the geometries found for six-coordinate Mentaa (n = 4 or 8) complexes. The two possibilities are cis-trigonal prismatic as exemplified by [Zr(Me4taa)Cl2] [Fig. 4(b)],4 and trans-octahedral as found for [Sn(Me4taa)Cl2] [Fig. 4(c)].5a Because of the relatively small hole size and typical saddle-shape of the Mentaa ligands the cis-trigonal prism is the most commonly encountered of the two six-coordinate geometries. Note that in [Sn(Me4taa)Cl2] the Sn–Cl bond pointing ‘down’ (as drawn) is 0.068(2) Å longer than the other. This relative lengthening of the ‘down’ metal–ligand bond is a typical feature of transoctahedral Mentaa derivatives.Fig. 4 Typical Mentaa saddle-shaped conformations and metal coordination geometries. Solid state structures of (a) square base pyramidal [Fe- (Me4taa)Cl],2 (b) trigonal prismatic [Zr(Me4taa)Cl2],4 and (c) octahedral [Sn(Me4taa)Cl2].5a Hydrogen atoms are omitted for clarity and the remaining atoms are drawn as spheres of arbitary radius. 2.3 Theoretical studies A number of computational studies of dibenzotetraaza[14]annulene complexes have been carried out at various levels of sophistication (extended-H�uckel density functional SCF-Xa- SW) and in some instances the results have been augmented by photoelectron spectroscopic data. The key conclusions from the various studies are summarised below and underline some of the general points made above.2,6,7–10 To simplify computational matters it is possible to model the 6H4 groups of Me4taa with H and cis-C2H2 methyl and o-C respectively.6,7 That o-C6H4 can be successfully substituted by cis-C2H2 emphasises the lack of conjugation between the 1,3-diiminato and o-phenylene fragments which was confirmed by density functional calculations for the whole complex [Ni(taa)].8 Furthermore calculations for the free planar and non-planar dibenzotetraaza[14]annulene ligands,9 and also for the metal complexes [Cu(taa)] (planar ligand) and [Cu(Me4taa)] (saddle-shaped ligand),8 found minimal electronic structure effects on the macrocycle donor orbitals despite the quite different ligand conformations.5H5)2]2+.6 As expected the dibenzotetraaza[14]annulene ligands carry a formal dinegative charge. This is localised mainly on the nitrogen atoms but a significant negative charge is also found on the so-called g-carbons [i.e. the ones bearing R1 substituents in Fig. 1(a)].9 In contrast the iminato (–CNN) carbons carry a partial positive charge. This charge distribution gives rise to electrophilic (at the g-carbons) and nucleophilic (at the iminato carbons) patterns of attack as described in Section 5 below. The ability of the iminato carbons to ‘carry’ some of the positive charge also for example contributes towards the reduced electrophilicity of the [Zr(Me4taa)]2+ fragment versus the apparently related [Zr(h-C For the later transition metal complexes [M(dibenzotetraaza[ 14]annulene)] (M = Cu Ni) metal–ligand p-bonding effects appear to be quite small although it is thought that the ligands can act as modest p-acceptors.Calculations for the binuclear complex [Cr2(Me4taa)2] however suggest that there is sufficient p-donation from the macrocycle to disrupt the formation of a possible Cr–Cr d bond.10a The dichromium complex thus possesses only a metal–metal triple bond despite the eclipsed geometry found in the solid state. In contrast the dimolybdenum congener [Mo2(Me4taa)2] is proposed to have a full Mo–Mo quadruple bond because the more diffuse 4d orbitals of Mo overlap better with the orbitals of the neighbouring metal than with the p-donor orbitals of the Me4taa.2.4 Choice of dibenzotetraaza[14]annulene ligand It is the R2-methylated (i.e. Mentaa-type) systems that have been most extensively studied. This is probably due to their unusual and interesting properties ease of synthesis and generally better solubility. For example (at the time of writing) there are currently over 140 crystallographically characterised dibenzotetraaza[ 14]annulene ligands and/or their complexes on the Cambridge Crystallographic Database but only about twenty of these have R2 = H. The examples given throughout this article reflect the emphasis on the Mentaa homologues but refererence will be made to the taa-type systems as appropriate. 3 Preparative methods 3.1 Ligand syntheses This section describes the general methods used to prepare dibenzotetraaza[14]annulenes.It has long been known that o-phenylene diamines and 1,3-dicarbonyl compounds (Scheme 1) undergo 1 1 condensation reactions to form 1,5-benzodiazepines. Therefore this type of direct approach is not suitable for the synthesis of macrocycles. However reliable synthetic routes to the dibenzotetraaza[14]annulenes are now well-established and can be divided into two distinct categories depending on the kind of R2 substituent required.2,11–15 R2 O O N NH2 R3 R3 R1 H+ + H R2 R2 –2 H2O N R3 R3 NH2 H R1 R2 Scheme 1 For R2 = H non-template methods (Scheme 2) may be used starting from o-phenylene diamines and either propynal or substituted acroleins as first reported by Hiller in 1968.11 The acrolein method is useful for introducing R1 = alkyl or aryl substituents into the g-position.For dibenzotetraaza[ 14]annulenes where R2 ­ H template methods (Scheme 3) are required. The template synthesis of tetramethyl dibenzotetraaza[ 14]annulene was first reported by J�ager in 1969 and subsequently developed by Goedken and others.2,14,15 Typically Ni2+ (introduced as the tetrahydrate of the diacetate salt) is used as the templating ion in a reaction between an o-phenylene diamine and a 1,3-dicarbonyl species. The metal can then be stripped from the macrocycle using anhydrous hydrogen 107 Chemical Society Reviews 1998 volume 27 H C C R3 R3 NH2 R3 2 R3 NH2 + O O 2 R2 R2 H R1 chloride to afford the desired macrocycle after a straightforward workup.Once the macrocycle has been formed further substitution of certain R1 functional groups into on (more usually) both of the g-positions is possible starting either from the neutral macrocycle itself or more typically from the metal complex (Scheme 4). The type of R1 group that may be introduced this way appears to be restricted to benzyl and benzoyl functionalities although these can contain quite elaborate substituents (such as crown ethers).16 3.2 Synthesis of metal complexes There are two principal methods (i.e. template and nontemplate) for preparing metal complexes of dibenzotetraaza[ 14]annulenes. Scheme 5 shows some specific preparations by way of example. As already seen (Scheme 3) template syntheses for Ni2+ (in some instances also for Co2+ and Cu2+) can be used to effect the direct cyclisation of o-phenylene Chemical Society Reviews 1998 volume 27 108 NH2 R3 R3 NH2 + O H O or H X H R1 R1 = alkyl aryl X = OEt NMe2 R1 H H R3 N N H H N N R3 H H R1 Scheme 2 R1 R2 R2 R3 N N R3 Ni2+ Ni reflux N N R3 R3 R2 R2 R1 (i) HCl (g) (ii) Neutralise R1 R2 R2 R3 N N R3 H H N N R3 R3 R2 R2 R1 Scheme 3 H R2 R2 R3 N N R3 [Y] N R3 R3 R2 R2 H N R1-X R1 R2 R2 R3 N N R3 [Y] N N R3 R3 R2 R2 R1 [Y] = Ni or (H)2 R1 = ArC(O)- ArCH2- (NC)2CNC(CO2Et)-; X = Cl or Br Scheme 4 diamines with 1,3-dicarbonyl compounds.Furthermore although metal ions are not essential for macrocycle formation from o-phenylene diamines with propynal or 3-substituted acroleins (Scheme 2) carrying out these reactions in the presence of Co2+ Ni2+ or Cu2+ directly affords the corresponding MII macrocyclic complexes [Scheme 5(a)].12a,13 The template syntheses described above are generally restricted to the later transition metals only.For other transition and main group metal derivatives routes based on the reaction of the preformed neutral macrocycle itself or its dilithiated derivative are used. The dilithium species Li2[Mentaa] are themselves conveniently prepared by treating the neutral dibenzotetraaza[14]annulene with methyl- or n-butyl-lithium in a hydrocarbon solvent.4 The dme (dme = 1,2-dimethoxyethane) adduct ‘Li2[Me4taa]·1.5(dme)’ has been crystallographically characterised (see below) but the solid state structures of the non-solvated species are unknown.The choice of using a neutral versus a dilithiated macrocycle generally depends on the metal substrate to be employed and the desired product. If for example metal alkyl or amide groups are to be substituted then using the neutral macrocyle will give simple alkane or amine by-products that are easily separated from the product complex [Scheme 5(b)]. On the other hand if metal halide substrates are to be used then the dilithiated ligand may be the better choice [Scheme 5(c)] since use of a neutral macrocycle generally requires the addition of a tertiary amine base to take up the hydrogen halide that will be eliminated.Because lithium halide side products are often more easy to separate than alkylammonium halides using the dilithiated ligand is generally more convenient. The best choice of reagent for other metal substrates (such as acetate or carbonyl complexes) is made according to similar considerations. 4 Unusual ligand conformations and coordination modes Section 2.2 and Fig. 4 summarised the typical conformations and coordination modes found for dibenzotetraaza[14]annulene ligands. This having been established it is appropriate now to consider some of the more interesting exceptions. 4.1 Ligand conformations Dibenzotetraaza[14]annulene complexes with R2 = H offer few surprises in this regard and are either planar or slightly saddleshaped.The R2 = Me derivatives however can show some (a) O 2 H C C H NH2 or + 2 NH2 O H 2 Me2N Ph (b) N N H 0.5 Al2R6 H H –2 RH N N (c) – N N (Li+)2 N N 3.14,15 – Scheme 5 Examples of the synthesis of dibenzotetraaza[14]annulene compexes using (a) template,11–13 and (b c) non-template methods.4,21 See also Scheme interesting and unusual conformations. As explained for the Mentaa (n = 4 or 8) systems the most typical conformation of the complexed ligand is analogous to that found in the solid state for the free H2Mentaa themselves—i.e. substantially non-planar with the o-phenylene rings tilted ‘up’ (i.e. towards the metal coordination site) and the 1,3-diiminato linkages oriented ‘down’ [see Figs.3(b) and 4]. Fig. 5 shows three examples of Mentaa ligands not adopting the conventional geometry. In Fig. 5(a) the compound [Ge(Me8taa)Te] has an inverted saddle-shaped ligand (i.e. with the o-phenylene rings ‘down’ and the 1,3-diiminato moieties ‘up’).17a Intriguingly the homologous sulfido and selenido complexes both have Me8taa ligands possessing the conventional configuration. In Fig. 5(b) and (c) the dibenzotetraaza[14]annulenes do not adopt any kind of saddle-shape at all and in these two centrosymmetric complexes the ligands are best described as ‘step-like’. In [Pd(Me4taa)] [Fig. 5(b)] the 1,3-diiminato linkages are tilted in opposite directions.18 The tilts are not very pronounced and therefore lead to a significant twist deformation of the o-phenylene rings in order to minimise steric interactions with the methyl groups.In the bis(triphenylphosphine)ruthenium example [Figs. 5(c)] it is the o- phenylene rings that are tilted in opposite directions and the 1,3-diiminato fragment now must undergo a twist deformation to avoid unfavourable steric interactions. The bis(methyldiphenylphosphine) homologue also exhibits similar features.19 It may be the favourable ligand field stabilisation of the second row PdII (low spin d8 favouring square planar coordination) and RuII (low spin d6 favouring octahedral coordination) complexes that helps drive the unusual distortion of the ligands so as to obtain the most favourable metal coordination geometry. R1 N N M2+ M N N R1 M = Co Ni Cu; R1 = H or Ph R N N Al N N R = Me or Et Cl Cl N N Zr ZrCl4(thf)2 N –2 LiCl 4taa)][AlCl4]2.20 2- N 4.2 Ligand coordination modes So far we have encountered only mononuclear dibenzotetraaza[ 14]annulene metal complexes in which the macrocycle acts as a doubly deprotonated dianionic tetradentate ligand [Fig.2(a)] to a single metal centre. Examples of this type of coordination are illustrated in Figs. 4 and 5. Although this is by far the most common situation found for the dibenzotetraaza[ 14]annulene complexes other coordination modes and metal nuclearities are possible as illustrated by Fig. 6. The complex [W(h2-H2Me4taa)(CO)4] [Fig. 6(a)] prepared from H2Me4taa and W(CO)6 has the macrocycle acting as a neutral bidentate ligand.2 Note that in this complex (and its Mo congener) the H2Me4taa exists as the unusual triimine monoeneamine tautomer [Fig.2(b)] as opposed to the diimine diamine form found in the free ligands [Fig. 1(a)]. A tetraimine H2Me4taa tautomer [Fig. 2(c)] has recently been structurally characterised for the neutral ligand in [Ni(H2- Me The main group derivative [Ga(h2-HMe4taa)Me2] [Fig. 6(b)] also has the macrocycle coordinated in a bidentate fashion but in this instance it acts as a monoanionic ligand with a single N–H group remaining [Fig. 2(d)].21 This coordination mode has also been found in [Mo(h2-HMe4taa)(O)2(acac)],22a and a bridging monoanionic bis-bidentate HMe4taa was recently reported for a diruthenium complex.22b The third example [Fig.6(c)] shows a dianionic Me4taa ligand bridging a Mo–Mo quadruple bond in a bis- bidentate coordination mode that was first reported by Goedken for the dirhodium species [Rh (Me4taa)(CO)4].2 The final example [Fig. 6(d)] shows the lithium ion local coordination geometry at one ‘end’ of the interesting tetralithium species [(dme)Li(m-Me4taa)Li(m-dme)- Chemical Society Reviews 1998 volume 27 109 Li(m-Me4taa)Li(dme)] (‘Li2[Me4taa]·1.5dme’).4 In this complex each dianionic Me4taa macrocycle acts as a bistetradentate ligand to a pair of Li+ ions located either side of the N4 plane. The remainder of the lithium coordination sphere is comprised of oxygen donors from the dme ligands. 5 Reactivity of the dibenzotetraaza[14]annulene ligands The majority of the reaction chemistry of dibenzotetraaza[ 14]annulene complexes concerns changes at the metal centre with the macrocycle acting only as a supporting ligand (see the following Sections).As already seen in Scheme 4 however reaction of certain complexes with benzyl or benzoyl halides can lead to facile R1 group substitution into the macrocycle at the g-positions. Indeed as Scheme 6 illustrates this is the most common site for reactions occurring at the ligand itself and Fig. 5 Unusual dibenzotetraaza[14]annulene ligand conformations. Solid state structures of (a) [Ge(Me4taa)Te],17a (b) [Pd(Me4taa)],18 and (c) [Ru(Me8taa)(PPh3)2].19 Hydrogen atoms are omitted for clarity and the Fig. 6 Unusual dibenzotetraaza[14]annulene coordination modes.Solid state structures of (a) [W(h2-H2Me4taa)(CO)4],2 (b) [Ga(h2- HMe4taa)- Me2],21 (c) [Mo2(Me4taa)2(OAc)2],2 and (d) the lithium local coordination geometry for one ‘end’ of [(dme)Li(m-Me4taa)Li(m-dme)Li(m-Me4taa)- Li(dme)].4 Most hydrogen atoms are omitted for clarity and the remaining atoms are drawn as spheres of arbitary radius. remaining atoms are drawn as spheres of arbitary radius. Chemical Society Reviews 1998 volume 27 110 N N ( a) (a) 2 Ni N N ( b)(b) py + N N H-C C-H Co N N ( c)(c) R R N N Zr N N reflects the activated character of the g-carbon in complexed dibenzotetraaza[14]annulenes. Thus Scheme 6(a) shows that electrochemical or chemical oxidation of [Ni(Me4taa)] does not give a stable radical cation and instead a dimeric compound containing two macrocyclic moieties linked by a new carbon–carbon bond is formed.23a This dication is easily deprotonated at the bridging C–H groups to form the corresponding neutral species.Similar oxidative couplings are found for Cu and Co complexes of Me4taa. Indeed electroactive surface-modified electrodes (featuring polymers of C–C linked metal macrocycles) may be prepared by the oxidative electropolymerisation of [Ni(Me4taa)] and its o-phenylene ring-substituted homologues.23b The C–C linking process is readily inhibited however by the presence of R1 ­ H substituents in the g-positions of the Me4taa ligand. In this case a stable ligand-centred radical cation is formed as is typical for unsaturated nitrogen-based macrocycles.4taa undergo C–C bond Certain Co and Rh complexes of Me forming reactions with alkynes nitriles and strained cyclic alkenes (see also Section 7) at the g-position to form new pentacoordinating macrocycles.2,24 Scheme 6(b) shows the reaction between [Co(Me4taa)(py)]+ and ethyne. Hoffmann has suggested that it is the negative charge build up (see Section 2.3) at the g-position that is responsible for the formation of this Scheme 6 Reactions involving coordinated dibenzotetraaza[14]annulene ligands.6,23,24 R = Me or CH2Ph 1,4-cycloadduct.9 The cationic bidentate macrocycle complexes [M(h2-HMe4taa)(CO)3(NO)]+ (M = Mo or W) also undergo C–C bond formation at the g-position.2 Thus reaction with MeCN gives rise to a new tri-coordinate ligand which may be viewed as a monoprotonated relative of those formed in the cobalt systems above.All the above examples involve reaction at the g-position. However high oxidation state early transition metal alkyl complexes such as [Zr(Me4taa)R2] (R = Me CH2Ph) can undergo intramolecular migratory nucleophilic attack at an iminato carbon atom.6 This reaction [Scheme 6(c)] forms a new trianionic macrocyclic ligand and is thought to be promoted by the high oxidation state metal centre generating an unfavourably large positive charge at the iminato carbon (see Section 2.3). Lower oxidation state metal [e.g. TiIII and VIII] alkyl complexes of Me4taa are however stable to this type of migration possibly as a consequence of the reduced positive charge at the iminato carbon atom.25 6 Recent highlights in main group chemistry In contrast to the wealth of transition metal chemistry of the dibenzotetraaza[14]annulenes (see also Section 7 below) there are still relatively few p-block metal compounds known and Chemical Society Reviews 1998 volume 27 N N Ni N N –2 e H H N N Ni N N py 2+ + N N Co N N H H H R N N Zr Benzene room remp.N N R 111 I N N Sn N N I Me E N N M N N M = Ge; E = S Se Te M = Sn; E = S Se OTf TfO Zr N N N N PhCH2MgCl R R N N M R3 R3 N N M = Zr Hf R3 = H Me; R = Me CH2Ph CH2SiMe3 ButNC (M = Zr R = CH2Ph) R R But But C C N N Zr N N N N R = CH2Ph Scheme 8 Selected group 4 organometallic and related chemistry.4,6,25,27–29 See also Scheme 6(c).6 Chemical Society Reviews 1998 volume 27 112 N N R3 R3 M I2 M = Sn R3 N N R3 M = Ge Sn; R3 = H or Me E or Te/PMe3 + IE N N M MeI N N M = Ge; E = S Se Te M = Sn; E = S Se S M = Ge; E = Se Te S N N Ge N N Scheme 7 Reactions of [M(Mentaa)] (n = 4 or 8; M = Ge or Sn).5,17 Cl Cl N N M R3 R3 AgOTf (M = Zr) R3 R3 N N M = Zr Hf; R3 = H Me LiR or R N R3 N Me M HNMe2Ph+ R3 Me N N M = Zr Hf; R = Me CH2Ph CH2SiMe3 Me3SiCºCMe (M = Hf R = Me) Me Me Me3Si N N Hf Me Me N N OC OC Cr(CO)5(THF) Al2Me6 S M = Sn; E = S Se Mg(C4H6)(thf)2 (M = Zr) + Me Me MeCºCMe (M = Hf R = Me) Me N Me + Me N Me Me CO Cr CO CO N N M N N M = Ge or Sn Me N N Al N N S E N N Sn N N Zr N N N N + Me Me Me Me N Me Hf Me N only two well-defined examples of s-block derivatives exist.The first structurally characterised main group derivative namely [Al(Me4taa)Et] was reported by Goedken and coworkers in 1984.21 The remainder of this Section gives key examples of the main group chemistry of dibenzotetraaza[14]annulenes. To date only Mentaa (n = 4 or 8) systems have been reported. For the electropositive metals Li Mg Al and Ga macrocyclic complexes are easily formed from the neutral H2Mentaa and the appropriate metal alkyl species as discussed in Section 3.2 for Li2[Me4taa] and illustrated in Scheme 5(b) for the synthesis of [Al(Me4taa)R] (R = Me Et).Similarly reaction of [Mg(CH2Ph)2] with H2Me4taa in thf affords the crystallographically characterised [Mg(Me4taa)(thf)] which has the usual saddle-shaped macrocycle with the Mg displaced from the N4 plane.26 The synthesis of [M(Mentaa)R] (M = Al Ga; R = Me or Et) proceeds via a surprisingly stable bis(alkyl) complex [M(HMentaa) R2] which has been structurally characterised for M = Ga n = 4 R = Me [Fig. 6(b)].21 Certain complexes [M(HMe4- taa)R2] only eliminate alkane on heating to above 100 °C (M = Al R = Et) or 200 °C (M = Ga R = Me) in the solid state. However reaction of H2Me4taa with Al2Me6 at 278 °C gives [Al(Me4taa)Me] directly in 95% yield.Dialkyl derivatives [M(Me4taa)Me2] (M = Si or Sn) are also known but are prepared from Li2[Me4taa] and the correponding MCl2Me2.5b Crystal structures for these complexes have not been reported but they are proposed to possess cis-dimethyl geometries. Metal halide and nitrate derivatives [M(Mentaa)(X)m] (for M = Ga or In m = 1; for M = Si Ge or Sn m = 2 n = 4 or 8; X = F Cl Br I or NO3) have been prepared and structurally Fig. 7 Solid state structures of (a) [Ge(Me4taa)] and (b) [(Me4- taa)GeCr(CO)5].17b Hydrogen atoms are omitted for clarity and the remaining atoms are drawn as spheres of arbitrary radius. characterised.5 The syntheses are generally via reaction of the appropriate metal halide with Li2[Me4taa] or H2Me4taa in the presence of NEt3 (but see also Scheme 7).The X-ray structures of [Sn(Me4taa)Cl2] [Fig. 4(c)] and [Sn(Me8taa)I2] reveal octahedrally coordinated Sn atoms with halide ligands in mutually trans positions.5a Main group derivatives with no additional ancilliary ligands exhibit interesting chemistry (Scheme 7). Thus reaction of Li2[Mentaa] with GeCl2·dioxane affords the crystallographically characterised [Ge(Mentaa)] {n = 4 [Fig. 7(a)] or 8 [Scheme 7]}. Similar procedures gave the Sn and Pb congeners.17 The complexes [M(Me towards electrophiles but as would be expected show different behaviour depending on the identity of M. Thus reaction with [Cr(CO) taa)MCr(CO5] for M = Sn or Ge [X-ray structure shown in Fig. 7(b). However with [Pb(Me4taa)] a facile transmetallation ntaa)] (M = Ge Sn Pb) are very reactive 5(thf)] gave the unusual bimetallic complexes [(Me4- reaction occurred to give [Cr(Me4taa)(CO)3] and ‘PbO’.All 4taa)] (M = Ge Sn Pb) complexes undergo 2Me6 to form [Al(Me4- three [M(Me transmetallation reactions with Al taa)Me] and unknown group 14 alkyl side products. The divalent complexes [M(Mentaa)] (M = Ge Sn) undergo oxidative addition reactions with certain alkyl halides halogens and elemental chalcogens (Scheme 7).17a Of particular interest are the metal–ligand multiply-bonded derivatives [Ge(Me8taa)- E] (E = Se Te) which were the first such terminal monoselenido or -tellurido complexes of Ge. Comparative reactivity studies of the [M(Me8taa)E] (M = Ge E = S Se Te; M = Sn E = S Se) systems have highlighted interesting similarities and differences in the reactivities of the terminal chalcogenido linkages as a function of the group 14 element.X-Ray diffraction analysis of the Ge and Sn complexes revealed significant metal–chalcogen multiple bond character described as a combination of M+–E2 M2NE+ and M2·E+ resonance structures. 7 Recent highlights in transition metal chemistry The chemistry of transition metal complexes of dibenzotetraaza[ 14]annulenes spans nearly thirty years and comprises the majority of research activity for these ligands. It is not possible to give a detailed account of this chemistry here. Cotton and Czuchajowska’s general overview of transition metal–Me4taa complexes appeared in 1990 and is currently the principal cited reference for this area.Readers should consult this review for a more detailed account of mono- and bi-nuclear transition metal Me4taa complexes up to that time.2 Sections 2 to 5 of this Chem. Soc. Rev. provided a general overview of the types of transition metal complex that may be formed and also described some recent examples. This current section will highlight current developments in the chemistry and new directions of research being pursued using dibenzotetraaza[14]annulene ligands. ntaa dianionic tetraazamacrocycles have received The Me much recent attention as potential alternatives to the ubiquitous bis(h-cyclopentadienyl) ligand set. The attraction of the Mentaa ligands in this regard is derived in part from the typical cisdisposition of the ancilliary X-ligands in complexes of the type [M(Mentaa)(X)2] (X = halide hydrocarbyl or related; M = group 4 or 5 metal).This is also the situation found for [M(h- 5R5)2(X)2] and is known to be an essential characteristic of C alkene polymerisation catalysts. The organometallic and coordination chemistry of early ntaa complexes has been extensively transition metal Me developed by the groups of Floriani and Jordan in particular. 4,6,25,27–29 Scheme 8 shows examples of the synthesis and reactivity of group 4 Mentaa-supported group inorganic and organometallic complexes. As shown in [Scheme 6(c)] [Zr- (Me4taa)(R)2] (R = Me or CH2Ph) are unstable to thermallyinduced intramolecular migration of an R-group to one of the iminato carbon atoms.Cationic d0 complexes of the type [M(Me8taa)(R)]+ (prepared using routes previously developed 113 Chemical Society Reviews 1998 volume 27 Me (Li+)2 Me Zr(NAr)Cl2(py)3 Ar N N N Ar = 2,6-C Ar-N N N M = Ti Ar = Ph R = Tol M = Zr Ar = 2,6-C6H3Pri 2 R = But for metallocene systems) and these undergo C–H bond activation and insertion reactions.27 However the [Zr(Me8- taa)(h2-CH2Ph)][B(C6F5)4] system (either with or without added Al(Bui 3) cocatalyst) exhibits only low ethene polymerisation activity compared to that of the corresponding [Zr(h- C5H5)2Me2]/[HNMe2Ph][B(C6F5)4] system which is about 100 times more active under identical conditions. In a similar vein the [M(Me8taa)(R)]+ systems are less reactive for alkyne insertion than are [M(h-C5H5)2R2]+ species.The M–R and M–NMe2 bonds in the complexes [M(Me8- taa)(X)2] (X = alkyl or NMe2) are highly polarised as indicated by their reaction with CH2Cl2 to yield [M(Me8taa)Cl2] and RH or CH2(NMe2)2 respectively and by the reaction of [Zr(Me- 8taa)(NMe2)2] with Al2Me6 to yield the interesting Zr/Al binuclear complex [Zr(Me8taa)(m-NMe2)2AlMe2][AlMe4].27 This enhanced nucleophilicity of the Me8taa-supported systems (with respect to that of the metallocene analogues) apparently reflects the harder character of the macrocyclic ligands. In accordance with the above Floriani and coworkers have shown that the bis(triflate) complexes [Zr(Me4taa)(OTf)2] (Tf = CF3SO2) act as homogenous Lewis acid catalysts and the best results were obtained for the Mukaiyama aldol reaction and metal-assisted allylations.28 It was proposed that the bis(triflate) complexes mask the active mono- or di-cationic forms [Zr- (Me4taa)(OTf)]+ or [Zr(Me4taa)]2+ respectively which are thought to be generated in solution because of the typically labile nature of the triflate ligand.Dibenzotetraaza[14]annulenes have also provided a suitable environment for the study of early transition metal–chalcogenido and –imido multiple bonds (Scheme 9).2,22a30–33 The oxo and sulfido complexes [M(Me Chemical Society Reviews 1998 volume 27 Scheme 9 Synthesis and selected reactions of early transition metal–ligand multiply-bonded complexes,2,22a,30–33 4taa)(E)] (M = Ti or V; E = O or 114 Cl Cl Me N N M N N R3 R3 N N N N Me M = Ti V; R3 = H Me 3 RNH2 Ti(NBut)Cl2(py)3 (M = V) R py E N Zr N N M N N N 6H3Pri 2 ButNCO M = Ti R =But M = V; R = alkyl aryl or NMe2 PhNH2 (M = Ti) O C Ph N-R N M TolNCO N N Ti N N N N E (ButMe2Si)2E N N M Me Me R3 Me Me N R3 N Me3SiCl ( M=Ti) M = Ti V; E = Se Te O2 or N2O (M = Ti) H2E (M = Ti) E N N M R3 R3 H2E M = Ti R = But R3 R3 N N M = Ti or V E = O or S R3 = H or Me Re(Cp*)(CO)2(NO)+ + OC Cp* Re ON C O O N Ti N N N S) were first reported in the early 1980s by Goedken and Ladd.2 However the chemistry of these complexes was not developed until ten years later.Furthermore the first group 4 and 5 terminal imido hydrazido selenido and tellurido Mentaa derivatives have only very recently been described.As with the metal–ligand multiply-bonded main group complexes [M(Me8- taa)(E)] discussed in Section 6 these macrocyclic ligands allow the isolation and study of comparatively reactive unsaturated linkages. In particular the metal–ligand multiple bond in the oxo and imido complexes [Ti(Me4taa)(E)] (E = O NBut or NAr) have a well-established reaction chemistry at the Ti = E functional group including cycloaddition reactions. In the area of bioinorganic chemistry dibenzotetraaza[ 14]annulene complexes of the middle to late transition metals have continued to be studied as models for the active sites of certain naturally-occurring enzymes.The complex [Co(Me because of the resemblance of the macrocyclic ligand to the corrin ring of vitamin B of that shown in Scheme 6(b) addition of the strained alkene 4taa)] has been studied by Moore and coworkers 12 coenzyme.24 In a reaction reminiscent norbornene to [Co(Me4taa)] followed by air oxidation afforded 4taa g-carbon atoms. The a cationic complex in which the CNC bond of norbornene had added across the Co and one of the Me crystallographically characterised product contains a pendant alkyl–cobalt bond and may be viewed as a new model compound for vitamin B12 coenzyme. An analogous reaction occurs between [Rh2(Me4taa)2] and norbornene. Other biomimetic systems have also been reported recently. The reductive dehalogenation of alkyl halides by Na[BH4] or Na[BH(OMe)3] to yield alkane is efficiently catalysed by [Ni(Me4taa)].34 This system is considered to be a mimic of factor F430 the active centre of methyl coenzyme M reductase.Manganese(iii) complexes of peripherally-substituted dibenzotetraaza[ 14]annulenes are effective catalysts for the oxidation of certain nitroso compounds to the nitro derivatives and in this respect may be considered mimics of the function of the cytochrome P-450 enzymes.16a In another area of activity electrochemical studies of middle to late transition metal dibenzotetraaza[14]annulene complexes can lead to the synthesis of new materials.23,35 Modification of the ring substituents can affect the redox properties of the complexes and influence their tendency to form dimeric or polymeric systems via ring-coupling reactions (see Section 5).Further electrochemical studies of mono- and bi-nuclear dibenzotetraaza[14]annulene complexes have recently been reported. Electrochemically-generated copolymers of [Ni- (Me4taa)] and pyrrole have been prepared from solutions containing carefully-controlled amounts of the two monomers. 8 Concluding remarks In this brief overview it has been shown that dibenzotetraaza[ 14]annulenes are the organometallic and coordination chemist’s flexible friends. These ligands are tunable easily prepared and readily introduced. They provide a reliable and generally robust framework that continues to allow new main group and transition metal chemistry to flourish.9 Acknowledgements My thanks are extended to Professors Cotton Floriani Jordan Parkin and Wallbridge and to Drs Blake Davies Kuchta and McInnes for their helpful comments on the manuscript. I also acknowledge use of the EPSRC’s Chemical Database Service at CCCLRC Daresbury Laboratory. 10 References 1 L. F. Lindoy The Chemistry of Macrocyclic Ligand Complexes Cambridge University Press Cambridge 1989. 2 For a review of Me4taa and its transition metal complexes up to about 1990 see F. A. Cotton and J. Czuchajowska Polyhedron 1990 9 2553 and references therein. 3 E. Sister V. Gottfried M. Kapon M. Kaftory Z. Dori and H. B. Gray Inorg. Chem. 1988 27 600. 4 S. de Angelis E. Solari E. Gallo C. Floriani A. Chiesi-Villa and C.Rizzoli Inorg. Chem. 1992 31 2520. 5 (a) M. C. Kuchta and G. Parkin Polyhedron 1996 15 4599 and references therein; (b) G. R. Willey and M. D. Rudd Polyhedron 1992 11 2805. 6 L. Giannini E. Solari S. De Angelis T. R. Ward C. Floriani A. Chiesi- Villa and C. Rizzoli J. Am. Chem. Soc. 1995 117 5801 and references therein. 7 J. M. Kerbaol J. E. Furet J. E. Guerchais Y. Le Mest J. Y. Saillard J. Sala-Pala and L. Toupet Inorg. Chem. 1993 32 713 and references therein. 8 G. Ricciardi A. Bavoso A. Rosa F. Lelj and Y. Cizov J. Chem. Soc. Dalton Trans. 1995 2385 and references therein. 9 K. Tatsumi and R. Hoffmann Inorg. Chem. 1981 20 3771. 10 (a) F. A. Cotton J. Czuchajowska and X. Feng Inorg. Chem. 1990 29 4329; (b) F. A. Cotton J. Czuchajowska and X.Feng Inorg. Chem. 1991 30 349. 11 C. L. Honeybourne and P. Burchill Inorg. Synth. 1978 18 44 and references therein. 12 P. J. Lukes A. C. McGregor T. Clifford and J. A. Crayston Inorg. Chem. 1992 31 4697 and references therein. 13 J. Jubb L. F. Larkworthy D. C. Povey and G. W. Smith Polyhedron 1993 12 1179. 14 V. L. Goedken and M. C. Weiss Inorg. Synth. 1980 20 115 and references therein. 15 F. A. L’Eplattenier and A. Pugin Helv. Chim. Acta 1975 58 917. 16 (a) J. Eilmes Polyhedron 1992 11 581 and references therein; (b) K. Sakata M. Shimoda and M. Hashimoto J. Heterocyc. Chem. 1996 33 1593 and references therein. 17 (a) M. C. Kuchta and G. Parkin Chem. Commun. 1996 1669 and references therein; (b) D. A. Atwood V. O. Atwood A.H. Cowley H. R. Gobran and J. L. Atwood Inorg. Chem. 1993 32 4671 and references therein. 18 M. Tsutsui R. L. Bobsein G. Cash and R. Pettersen Inorg. Chem. 1979 18 759. 19 L. Luo E. D. Stevens and S. P. Nolan Inorg. Chem. 1996 35 252 and references therein. 20 Y.-H. Tsai and C.-C. Lin Chem. Lett. 1996 101. 21 J. C. Cannadine W. Errington P. Moore M. G. H. Wallbridge E. Nield and D. Fenn J. Organomet. Chem. 1995 486 237 and references therein. 22 (a) S. Lee C. Floriani A. Chiesi-Villa and C. Guastini J. Chem. Soc. Dalton Trans. 1989 145; (b) F. A. Cotton and J. Czuchajowska J. Am. Chem. Soc. 1991 113 3427. 23 (a) J. C. Dabrowiak D. P. Fisher F. C. McElroy and D. J. Macero Inorg. Chem. 1979 18 2304 and references therein; (b) C. L. Bailey R. D. Bereman and D. P. Rillema Inorg. Chem. 1986 25 3149 and references therein. 24 P. Kofod P. Moore and N. W. Alcock Acta Chem. Scand. 1993 47 1083 and references therein. 25 E. Solari S. De Angelis C. Floriani A. Chiesi-Villa and C. Rizzoli Inorg. Chem. 1992 31 96 and references therein. 26 F. Corazza C. Floriani A. Chiesi-Villa C. Guastini and S. Ciurli J. Chem. Soc. Dalton Trans. 1988 2341. 27 A. Martin R. Urhammer T. G. Gardner R. F. Jordan and R. D. Rogers Organometallics 1998 17 382 and references therein. 28 P. G. Cozzi C. Floriani A. Chiesi-Villa and C. Rizzoli Synlett 1994 857. 29 L. Giannini E. Solari C. Floriani A. Chiesi-Villa and C. Rizzoli Angew. Chem. Int. Ed. Engl. 1994 33 2204. 30 P. Mountford Chem. Commun. 1997 (feature article) 2127 (also available at http://www.rsc.org/ccenhanced) and references therein. 31 J. L. Kisko T. Hascall and G. Parkin J. Am. Chem. Soc. 1997 119 7609 and references therein. 32 H. Schumann Inorg. Chem. 1996 35 1808 and references therein. 33 C. E. Housemekerides D. L. Ramage C. M. Kretz J. T. Shontz R. S. Pilato G. L. Geoffroy A. L. Rheingold and B. S. Haggerty Inorg. Chem. 1992 31 4453 and references therein. 34 T. Arai K. Kimio H. Kondo and S. Sakaki Bull. Chem. Soc. Jpn 1994 67 705. 35 (a) B. Keita Y. W. Lu and J. Napjo J. Electroanal. Chem. 1994 367 285 and references therein; (b) A. Deronzier and M.-J. Marques J. Electroanal. Chem. 1994 370 151 and references therein. Received 28th July 1997 Accepted 5th November 1997 115 Chemical Society Reviews 1998 volume 27

ISSN:0306-0012    

DOI:10.1039/a827105z

出版商:RSC    

年代:1998

数据来源: RSC

摘要:

Liquid–liquid equilibria in polymer solutions at negative pressure Attila Imrea and W. Alexander Van Hookb a Physical Chemistry Department Atomic Energy Research Institute POB 49 H-1525 Budapest Hungary b Chemistry Department University of Tennessee Knoxville TN 37996 USA Properties of liquids under tension (i.e. at negative pressure) are discussed together with methods of producing negative pressure. That established the pressure dependence of liquid–liquid demixing in certain polymer–solvent solutions including demixing at negative pressure is described. Yes we know that many physical quantities cannot be negative by their very nature. There is no negative volume no negative mass and early on probably in High School we learned that temperature once properly expressed using the Kelvin (absolute) scale has a natural zero.It is intrinsically positive. (At least this is true for systems in thermodynamic equilibrium; quantum systems forced far from equilibrium— LASERS MASERS etc.— show population inversions which can be usefully described using the idea of negative temperature.1 Ordinarily however with feet firmly on classical ground one thinks of temperature as inherently positive.) And what about pressure? In beginning science courses pressure is usually introduced during a discussion of introductory level kinetic-molecular theory. In this approach pressure which is rigorously and simply defined as applied force per unit area is derived by calculating the net momentum change (force) per unit area which results when particles hit a wall then rebound.At that level of analysis pressure is certainly a positive number. As particle density drops so does pressure and this is satisfying. At zero number density (i.e. in a vacuum) one calculates zero pressure thereby falling into the logical trap of (correctly) correlating pressure with number density for the example under discussion then later on (incorrectly) generally identifying or implicitly defining pressure as proportional to W. Alexander Van Hook was born in 1936 and received his education at the College of the Holy Cross (BS 1957) and Johns Hopkins University (MA 1960 PhD 1961 mentor P. H. Emmett). After a postdoctoral stay at Brookhaven National Laboratory spent with Jacob Bigeleisen and Max Wolfsberg he joined the University of Tennessee where he is presently Ziegler Professor of Chemistry.Van Hook spent 1967–1968 at Universite Libre de Bruxelles studying with I. Prigogine and G. Thomaes and a good deal of 1972 at the Boris Kidric Institute Belgrade. Shorter study visits have been spent at Lanzhou and Beijing Universities PRC and in Budapest and Lisbon. Van Hook’s research interests are in isotope effects on the properties of condensed phases solution thermodynamics and light and neutron scattering of polymer–solvent systems. W. Alexander Van Hook number density. We know this line of thought is correct for gases in the low density limit (ideal gases) and by implication think it should work for liquids and solids.It is for reasons like this that most of us (even engineers chemists and physicists) think of pressure as always positive. Actually this is wrong. Pressure is not necessarily positive. It is simply and rigorously defined as force per unit area—and force is a vectorial quantity. One can pull as well as push—we push on an object to pressurize it—by pulling we must depressurize. If we pull hard enough we place the object under tension and it is useful to articulate this result as corresponding to negative pressure. One cannot do this for gases—they collapse—and that accounts for the logical trap of the last paragraph. One can do it for solids they are strong enough to resist tensile forces tending to distortion perhaps to very high tensions—at which point they fracture.Even so although it is easy to pull on a solid (at least in one or another of its dimensions) it is harder to imagine similar experiments on fluids. The problem is how to support the fluid structure and prevent its fracture as tension increases i.e. as pressure falls below zero. Hard but not impossible. In 1662 only 19 years after Torricelli’s studies and the development of the barometer Christian Huygens experimented with a long Toricelli-tube which he completely filled with mercury and then inverted. To his surprise he found that almost 2 m of mercury could be supported in the tube and this without increasing the outer pressure (provided the mercury was carefully degassed).2 This is shown in Fig. 1 which compares two Torricelli tubes.Both have been filled with carefully degassed mercury and then Attila Imre was born in 1966. He graduated from the Physics Department of E�otv�os University Budapest in 1990 and received his PhD under the supervision of Dr T. Pajkossy (KFKI Atomic Energy Research Institute). During 1993–1996 he spent 22 1 years collaborating with Alexander Van Hook at the University of Tennessee. At present he is a senior researcher in the Physical Chemistry Department at KFKI Atomic Energy Research Institute. His research interests lie in the description of solid–liquid interfaces in polymer solubility and the properties of liquids under tension (including cavitation). Attila Imre 117 Chemical Society Reviews 1998 volume 27 inverted and placed in a dish of mercury.In high school science we learned that the atmosphere at sea level (on a nice day) supports a column of mercury 76 cm high [Fig. 1 left]. The space above the mercury in the lefthand tube (a) is occupied by mercury vapor (the famous Torricellian vacuum) and Pa = Ptorr = Pexernal 2 rHgghAB = 0.5 pa or 5 3 1026 atm. Here r is the density of mercury hAB is the height of the column (76 cm when Pext = 1 atm) and g is the gravitational constant. Huygens however carried out a slightly different experiment. He used carefully degassed mercury and most likely scrupulously cleaned and dried the glass tube prior to filling. This time [Fig. 1(b)] a much longer column of mercury was supported but clearly only 76 cm can be supported by the external pressure.External pressure must be independent of the particular Torricelli tube we use to measure it. One can calculate the pressure at the top of tube (b) just as we did for tube (a). For example should hAC = 152 cm the pressure at the top of tube (b) must be Pb = Pa2rHgghAC = 2101 325 pa = 21 atm. The liquid inside the tube is under tension; it is at negative pressure. The liquid column is supported in part by the external pressure and in part by the adhesive forces which exist between glass and mercury. Fig. 1 The tension manometer described a manometric device for generating negative pressure. (a) Torricellian barometer. A glass tube is filled with mercury then inverted and held in a dish of mercury as shown.In this manner the external barometric pressure (sea level in fair weather) is shown to be equivalent to a column of mercury 76 cm high. Minor corrections to account for variation in room temperature and capillary effects are required for high precision measurements. (b) The Huygens– Torricelli method for generating negative pressure. A glass tube is filled with mercury inverted and held in the dish as shown. At normal atmospheric conditions a column longer than 76 cm is supported partly by the atmospheric pressure and partly by adhesive forces. This ‘extra’ column of length BC is a measure of the negative pressure or tension on the liquid. See text for further discussion. (c) Schematic representation of data from the modified Huygens–Torricelli manometer.Almost two hundred years after Huygens the phenomenon shown in Fig. 1 was rediscovered by Donny (1848),3 then a little later by Berthelot (1850),4 although his experiment was a bit different—see below and finally by Reynolds (1882),5 all independently. In contrast to Huygens who did not have a correct explanation for the effect he witnessed these latter authors recognized the close ection between negative pressure and the adhesive forces between glass and liquid.2 This was clearly articulated by Donny who wrote in 1846:3 ‘It is well known to manufacturers of barometers that after the mercury has been boiled in one of these instruments when the device is slowly turned upright without tapping sometimes it happens that the mercury stays totally supported and only falls to its level relative to the weight of the atmosphere when the barometer is given a tap.However no Chemical Society Reviews 1998 volume 27 118 one has tried to get at the cause of this phenomenon nor has deduced its consequences. Only in the work of Laplace on capillary action do I find a passage where it seems the author has at least suspected a relation between this hanging of the mercury column and cohesion.’ Although one would think such multiple discoveries buttressed by the prominence of at least three of the four personalities would have brought this phenomenon forward to general recognition such has not been the case. The concept of negative pressure has unfortunately remained obscure. In this paper we join others2,6–8 in taking another small step in the direction of removing this obscurity.Our method will be to review some of our experimental observations on the pressure dependence of solubility of polymers including solubility at negative pressure. 1 Introduction 1.1 Polymer phase equilibria and negative pressure Many polymers dissolve in one or more well known solvents at all accessible temperatures (i.e. between the melting and critical points of the solvent) no matter how long the polymer chain. These are the so called “good solvents” and solutions of polymers in good solvents while viscous and perhaps hard to handle are nicely homogeneous [e.g. polystyrene (PS) in tetrahydrofuran (THF)]. Other solvents (e.g. PS–cyclohexane) only dissolve an infinite polymer chain between two well defined temperatures.Above and below those temperatures depending on the concentration we find precipitation followed by macroscopic phase separation into two fluid phases (one polymer rich the other polymer poor—therefore this is a liquid– liquid phase transition). In other cases the precipitating phase may be a solid. Our present interest is in the liquid–liquid case. The solvents in this case are known as q-solvents and the limiting temperatures (i.e. the precipitation temperatures of the infinite chain at its critical concentration which is found at or near the maximum in the solubility curve vide infra) are the socalled q-temperatures. Finally there exists a class of poor solvents which are unable to dissolve long polymer chains (and in some cases unable to dissolve even short chains) at appreciable concentrations.A good example is the PS–acetone system. Acetone will dissolve short chain versions of polystyrene but the limit (192 monomer units at the critical concentration) is low enough to destroy the utility of this solvent in all but special cases. So far in this section we have not spoken of pressure. The effect of pressure on solubility is well known especially in gas– liquid systems (Henry’s Law) but also in liquid–liquid systems. In polymer solutions pressure markedly affects solubility and the thermodynamic and molecular origins of that observation are now beginning to be understood. For polymers in q-solvents pressure usually increases solubility i.e.the solution remains homogeneous to lower temperatures than before and stays homogeneous to higher ones [Fig. 2(a) with the variable X equal or proportional to P]. Also poor solvents usually become better with the application of pressure so the extent of the homogeneous part of the phase diagram increases with P [Fig. 2(b) X equal or proportional to P] as does the limiting chain length which can be dissolved by that particular solvent.9 Let’s spend a bit more time discussing solutions in q-solvents and poor solvents. Fig. (2a) and (b) show phase diagrams for such solutions in (y,T,X) space [y = concentration (segment fraction) T = temperature X a third variable of interest]. For the moment consider a solution held at constant pressure (nominally 1 atm) and let X scale as a function of molecular mass (MW).The standard Flory–Huggins theory of polymer thermodynamics suggests X = MW21/2. In the figures X increases moving out from the page (MW dropping as X increases) and as expected the extent of the one phase homogeneous region expands with X. For the q-solvent [Fig. 2(a)] the locus of upper and lower consolute temperatures (in first approximation the critical locus) yields upper and lower curves in the (T,X)ycr projection with X = 0 intercepts (i.e. infinite MW) which define upper and lower q-temperatures. In Fig. 2(a) however the upper and lower curves are connected by a smoothing function (refer to the dotted line) extending into the hypothetical region X < 0.In Fig. 2(b) on the other hand we are in the poor solvent limit and the upper and lower branches join at a hypercritical point where X > 0 (i.e. at a real MW). In such an analysis one soon recognizes that the distinguishing difference between solutions in q-solvents and poor solvents is a shift in the phase diagrams along the X coordinate [compare Fig. 2(a) and (b)]. In the poor solvent case the (T,X)ycr Fig. 2 Schematic cloud point curves of polymer solutions. Liquid– liquid demixing for mixtures showing upper and lower consolute branches. The two phase regions are shaded. (a) For q-solvents with upper and lower q-temperatures. Temperature T is plotted against segment fraction polymer Y in the plane of the paper. The third variable might be either the molecular mass parameter X = MW21/2 at some chosen pressure or the pressure P at some chosen molecular mass.The curve drawn through the critical points and marked with heavy dots is further illustrated in Fig. 3. For the case X = MW21/2 qU and qL correspond to the (T X = 0) intercept of the line of ‘dots’ and the large open circle represents the hypercritical molecular mass parameter which has no physical meaning in this case but is quite useful in the mathematical description of the phase diagram. See text for further discussion. (b) For poor solvent systems showing upper and lower consolute curves joining at a hypercritical point (marked with the large solid circle). Axes labeled as in Fig. 2(a). The curve drawn through the critical points (heavy dots) is plotted in the (T,X) and (T,P) projections in Fig.3. See text for further discussion. M projection of the critical loci displays its extremum (hypercritical point) at real X i.e. X > 0. Below that point i.e. in the range (0 < X < Xhypercr) the system is collapsed into the hour glass configuration [see the darkest shading in Fig. 2(b)]. Solutions in q-solvents on the other hand have their extrema at X < 0 i.e. at negative pressure in the (T,y,X = P) W>0 projection (perhaps experimentally inaccessible) or negative MW21/2 in the (T,y,X = MW21/2)P>0 projection (definitely inaccessible). So much for projections at constant pressure. Now consider solutions at constant MW and set X equal to pressure. As P = X increases moving out from the plane of the paper (Fig.2) the solvent improves. In this state of affairs it should be possible to select solvent quality pressure temperature and MW such that the system of choice lies in the homogeneous region but not too far from either temperature or pressure induced phase transitions. From such a beginning one might induce precipitation by either raising or lowering the temperature or by lowering the pressure. Of course in a poor solvent one could also induce precipitation by increasing MW or by changing solvent quality but these are variables we earlier agreed to hold steady. At any rate one could argue that an even more judicious choice of solvent quality T and MW could place a solution in the homogeneous region at P ~ 0 but such that further lowering the pressure (to negative values placing the solution under tension) might induce precipitation.Of course this assumes that the equation of state describing the solution is well behaved and continuous across the boundary P = 0 and smoothly extends into the tensile region where P < 0. This is the kind of solution which we are going to focus on in the balance of this review. M ycrit,M The points made in the last few paragraphs are further illustrated in Fig. 3. Fig 3(a) is a (P,T,X = MW21/2)y constant concentration representation of the UCST/LCST demixing diagram. For y = ycrit the diagram includes (T,P)MW,ycrit sections at two MW values and (T,X = MW21/2)P,ycrit sections at four pressures including sections at both positive and negative pressures.The leftmost (T,P) W,ycrit projection is at X = MW21/2 = 0 and therefore maps the pressure dependences of the upper and lower q-temperatures. Fig. 3(b) shows two (P,T) W sections one (marked B) in a better solvent where the hypercritical demixing point is found at negative pressure the other (marked A) in a poorer solvent where it is found at P > 0. Later in this paper we will show an example of this type. The hypercritical points in Fig. 3(b) lie at the minima of the two curves and correspond to the points marked with encircled dots in Fig. 3(a). More than 20 years ago Wolf measured UCS and LCS loci for PS– diethyl ether solutions as a function of pressure.10 He observed that the UCS and LCS branches approached each other at low pressure and went on to speculate that they might join at even lower pressures (negative pressures) as in curve B Fig.3(b) but did not experimentally demonstrate that to be the case. In fact only recently has it been demonstrated in experiments from our laboratory that one can employ pressure and negative pressure to demonstrate continuity of state for these polymer solutions. These experiments have shown that certain systems with separate UCS and LCS branches join smoothly to yield a hypercritical point at negative pressure,11,12 and certain solutions in poor solvents which display a hypercritical point at P ~ 0 or P > 0 can literally be pushed into the q configuration by raising the pressure.12,13 To further illustrate we return to Fig.2. This time consider the two parts of the figure as (T,y,X = MW21/2)P projections taken at two pressures P[Fig. 2(a] > P[Fig. 2(b)]. It is the choice of solvent quality which dictates whether the hypercritical point lies at P > 0 P ~ 0 or P < 0 and in the material below we review experiments showing this to be the case. It must be understood in designing these kinds of experiments that one is strictly limited to negative pressures of magnitude smaller than the breaking strength (cavitation limit) of the liquid itself or the adhesive forces joining the liquid to the wall.6 119 Chemical Society Reviews 1998 volume 27 Fig. 3 (a) Schematics for UCST–LCST demixing in (T P X = MW21/2) space. Depending on the precise strength of the polymer–solvent interaction (i.e.on solvent quality) the system can make a transition from a q to a non-q (poor) solvent at positive pressure at a reachable negative pressure (tension) as shown or at a negative pressure too low to be experimentally observed (because obscured by cavitation). The dashed lines drawn through the dots show isopleths (T,P) projections at finite MW and Fig. 3(b) shows several such projections. The shaded areas are (T X = MW21/2) projections at several pressures both negative and positive. (b) (P,T)y,X projections of demixing isopleths for two solvent qualities. A = poor solvent B = qsolvent. We conclude this section by commenting that the ‘solubilityworsening’ effect of negative pressure has significant implications to begin with it demonstrates that the UCS and LCS precipitation branches are separate manifestations of the same phenomena.In other words the poor-solvent/q-solvent dichotomy is not rigid it can be shifted in either direction by pressure (and possibly by other variables). More practically this phenomenon gives us the opportunity to precipitate polymer from q-solvents simply by varying pressure without adding a third component (cosolvent). One could take advantage of this Chemical Society Reviews 1998 volume 27 120 in designing pressure fractionation processes for the purification of polymers. This might be useful for example in sharpening MW distributions. 1.2 Further remarks. Negative pressure in liquids and solutions Above we saw that the existence of negative pressure (tension) in liquids was experimentally confirmed more than three hundred years ago (see the historical descriptions given by Kell2 and Trevenna6).Stress is induced in a solid when it is pulled in the same sense negative pressures are generated in liquids when they are pulled. Of course one expects that negative pressure in liquids cannot be too deep or exist too long because liquids cannot hold stress for a long time but sometimes negative pressures (even tens of megapascals!) have been maintained for days,14 and theoretically (maybe) there is no time limit on moderate negative pressures.15 Certainly negative pressures of large enough magnitude may tear liquids just like big stresses fracture solids this maximum tension is the so-called homogeneous nucleation limit for cavitation.During the ‘three hundred year’ history of negative pressure research16 most effort lay in the attempt to reach the homogeneous nucleation limit.6 This limit is important in defining the equation of state for liquids. In addition there have been a few measurements on other physico-chemical properties. 17–19 Some researchers examined the effect of the negative pressure on the properties of various liquids including water and especially on the freezing point curves.20 However except for the liquid–liquid demixing experiments on PS solutions described by us,11,12 there has been very little work at negative pressure on the phase behavior of two component solutions.We are convinced that systematic measurements of physicochemical properties of liquids and solutions at negative pressure yield valuable information and that usefulness extends far beyond the mere mapping of the homogeneous nucleation limit. 2 Methods It is a well established maxim in hydraulics that pressurized fluids at rest are isotropically strained. Thus to make effective comparisons of fluid properties at positive and negative pressure it is important to ensure that the applied negative pressures are also isotropic. Anisotropic effects are common in polymer solutions under flow. For example the effects of shear on polymer solubility and on light scattering and neutron scattering from polymer solutions are well known21 and are of considerable theoretical and commercial interest.Nonetheless in the present context they are effects to be avoided. For the experiments described below we need a method to produce and maintain a large enough negative pressure preferably in a transparent cell (to permit direct observation of precipitation) for times which are long compared to the characteristic time of the phase transition. Additionally it will be convenient to provide mixing and it will be necessary to incorporate some method to measure or calculate the tension. Trevena6 describes several techniques to generate negative pressures in liquids. To begin with it is important to keep in mind that negative pressure is not a thermodynamically stable state i.e. any untoward disturbance or any contamination (especially suspended colloidal particles or traces of gases) can serve as nucleation centers to break the stressed liquid generating vapor bubbles (cavities) and initiating the process eqn.(1) whence the liquid pressure once more becomes (1) (metastable fluid)P < 0 ? (thermodynamically stable [fluid + vapor])P = P(vap) > 0 positive. Therefore any method used to produce negative pressures must either exclude or reduce these problems if one desires deep tensions. Methods in common use to place liquids under tension can be divided into two groups static and dynamic (see Table 1). By static we infer methods which produce and hold tensions for long times minutes to hours or even longer while dynamic methods only produce transient tensions which last but fractions of a second (typically a few milliseconds or less).Table 1 Some methods to generate negative pressures Dynamic methods Sudden pulling of a piston Backside of positive pressure wave Sound waves (acoustic method) Static methods Thermal pressure method. Tendency to contraction induced by cooling but prevented by liquid–wall adhesion. (Berthelot’s method or Vincent’s viscosity tonometer.27) Hydrostatic forces (Huygen’s method) Centrifugal force (Briggs’ method28) Flow through Venturi constriction Although static as viewed from (based on Bernoulli’s law). All dynamic methods basically employ a pressure pulse technique to generate tension.6 A pulse moving through a fluid generates a compressive (positive pressure) disturbance to its front and is followed by a decompression (negative pressure) to its rear.Pulses can be generated mechanically chemically (by explosions) or acoustically but in any of these cases the characteristic lifetime of the disturbance is determined by the mechanical relaxation time of the fluid which in turn is closely related to the speed of sound in that fluid. Thus the timescale during which the fluid at any point is under tension is of the order of milliseconds or less and the magnitude of that tension is rapidly changing even on the millisecond timescale. A second difficulty is that dynamically or acoustically generated tensile pulses are directional in character. This is very much a disadvantage since it is our goal to compare fluid properties of (isotropically) pressurized solutions with those same solutions under tension.Most of the static methods used to produce longer lasting ( > 1 s) tensions are based on the adhesion which exists between a solid wall and the fluid. Further information is found in Table 1. Trevena gives excellent and detailed discussions of the more common methods.6 Although the manometer method (Fig. 1) Comments Mechanical generation of short tension pulse of continuously changing value delocalized characterized by anisotropic pressure tensor Very short tensile pulse of continuously changing value delocalized anisotropic pressure tensor. Short cyclic tensile pulses of periodically changing value anisotropic localized when cell geometry permits standing waves.Comments Constant value of tension can be even for days (but not easily) isotropic pressure tensor requires very good temperature control. Tension changes continuously along the tube but is locally isotropic. High tensions require very long tubes and this is experimentally inconvenient. Used to generate very high tensions in the middle of a spinning capillary but the tension is anisotropic and changes continuously along the tube. an outer coordinate system this method is locally dynamic from the point of view of the liquid. Also the pressure tensor is anisotropic. first used by Huygens and briefly described early in this paper is conceptually simple its use to generate deep negative pressures in ordinary fluids (say 0.8 < r/(g cm23) < 1.2) would require capillaries many metres long.This is much too clumsy and the Huygens method is inconvenient for general use. The technique in most common use for generation of static fluid tensions and luckily the best for our purposes is the so called Berthelot-tube method4,6,14 (Fig.4). In the Berthelot method one almost fills a capillary with clean and degassed liquid or solution cools the sample now held under modest vacuum (thus eliminating residual air or other low boiling contaminants) then flame-seals the tube still under vacuum. At this point after the sample melts the capillary contains mostly fluid plus a small vapor bubble at the top [leftmost tube (tube 1 Fig.4)]. On heating the fluid expands so the vapor bubble disappears at a well-defined temperature which depends on the initial fluid/ vapor ratio and the thermal expansivity of the liquid [Tfill the filling temperature tube 2 Fig. 4]. Normally one continues to warm a few degrees above Tfill to dissolve all vapor bubbles and incipient nuclei and to ensure that liquid thoroughly permeates whatever pores exist at the wall (tube 3 Fig. 4). It is well established that pre-pressurizing almost always significantly increases the maximum tension in a given apparatus. That limit is usually determined by one or another of the mechanisms of heterogeneous nucleation of the vapor phase. In any event after heating (pre-pressurizing) the sample is cooled.As the temperature drops the fluid would contract were it not held to the wall by surface adhesive forces (tube 4 Fig. 4). The resulting tension accounts for the negative pressure. Obviously at Tfill p = pv where pv is the equilibrium vapor pressure of the fluid system. For T < Tfill the negative pressure can be calculated from the thermal properties of the liquid or solution using the thermal pressure coefficient G eqn. (2). (2) G = (¶P/¶T)V = 2 (¶V/¶T)P/(¶V/¶P)T = a/k Here a is the isobaric expansivity a = (1/V)(¶V/¶T)P = (¶lnV/ ¶T)P and k is the isothermal compressibility k = 2(1/V)(¶V/ ¶P)T = 2(¶lnV/¶P)T. Both a and k are weakly dependent functions of T and P so G = G(T,P). To first approximation we write eqn.(3) a = a (3) 0 + a1,T (T 2 Tfill) + a1,P (P 2 Pfill) + . higher order terms (H.O.T.) and eqn. (4) Fig. 4 The Berthelot method schematically illustrated. See text for detailed discussion. 121 Chemical Society Reviews 1998 volume 27 k = k0 + k1,T (T 2 Tfill) + k1,P (P 2 Pfill) + . H.O.T. (4) so eqn. (5) G = (a0/k0)[1 + (a1,T/a0 2 k1,T/k0)(T 2 Tfill) + 1,P/a0 2 k1,P/k0)(P 2 Pfill) + . H.O.T.] (5) (a Here a0 and k0 are the expansivity and compressibility at the fill temperature and pressure. For most fluids of interest the first order correction terms are small (i.e. a1,T/a0 k1,T/k0 a1,P/a0 and k1,P/k0 are all << 0). To sufficient precision one can thus write G = (¶P/¶T)V ~ G0 ~ (a0/k0). The negative pressure which results from cooling to a temperature some few degrees below Tfill is then eqn.(6) (6) ºdP ~ P 2 Pfill ~ P ~ º G0dT ~ G0(T 2 Tfill) As one continues to cool the Berthelot cell below T because |Pfill/P| < < 1. Typically fluids of interest have expansivities of the order of 1023 K21 and isothermal compressibilities of the order 1029 Pa21 so G0 ~ 106 Pa. K21 ~ 1 MPa K21. Also T < Tfill so P is negative and of the order of 21 MPa (10 bars) per degree of cooling. This simple calculation nicely demonstrates that the Berthelot technique is a simple and effective method to conveniently generate negative pressures of appreciable magnitude. Ordinarily G(T,P) is determined for P > 0 then extended to the fluid under tension. In an actual Berthelot experiment the pressure which is developed can be measured by determining the deformation of the container22 [which some Henderson and Speedy for example,20 construct in the form of a delicate spiral (an in situ Bourbon gauge)] or can be calculated from the temperature as outlined above.Alternatively it may be possible to incorporate a strain guage or other type pressure guage into the sample cell but this will be only at the risk of introducing new sites for heterogeneous nucleation. fill the pressure rapidly falls; before long the liquid breaks and tiny bubbles appear to witness the end of the negative pressure session (tube 5 Fig. 4). At this point the system jumps back to its equilibrium vapor pressure. Usually cavitation occurs at the wall–liquid interface and at a few MPa tension but Henderson and Speedy have succeeded in generating and maintaining measured negative pressures as high as ~ 20 MPa for highly purified water held in fine spiral capillaries,14,20 while Green et al.reached 80 MPa negative pressure (calculated from the thermal pressure coefficient) in water trapped in a quartz fissure.23 The homogeneous nucleation limit (i.e. that tension at which cavitation occurs because the cohesive forces between fluid molecules are exceeded by the applied tension) is usually unreachable in a practical sense. Below the heterogeneous limit usually seen at several MPa or more the tension can be held for minutes sometimes for days. The course of a phase transition induced by negative pressure in a polymer–solvent mixture can be seen in Fig.5. Point A represents the initial pressure and temperature—here the capillary contains liquid with small vapour bubbles and P = Pvapor. The filling temperature is at B. As the temperature is increased beyond B and towards C pressure increases rapidly. At the pre-pressurization limit C the sample is cooled the system passes through B again and thence into tension. With sufficient cooling the (T,P) path crosses the cloud point locus and the solution goes turbid (point D). Knowing the temperature difference between B and D (TD 2 Tfill) the tension can be calculated and recorded. One can stop the session by rewarming the system. Alternatively further cooling can cause cavitation (point E); in that case the pressure jumps discontinuously and rapidly clearing at point F and ending at G where P = Pfill = Pvapor.Each sample in its sealed capillary gives one point on the (T,P) isopleth for the cloud point. (The positive side of the cloud point curve can also be observed with this method,11 in that case adjusting the fill conditions to ensure the cp curve is located between points B and C.) Chemical Society Reviews 1998 volume 27 122 Fig. 5 The course of a Berthelot induced precipitation at negative pressure. The smooth curve is a (T,P) isopleth of the type illustrated in Fig. 3(b). The hypercritical point is located at negative pressure. The set of line segments AB–BC–CBDE–EFG describe the course of the Berthelot experiment to induce precipitation at negative pressure.See text for further and more detailed discussion. 3. Results two examples We have used the Berthelot negative pressure technique to study phase transitions in two separate polymer–solvent systems.24 In the first polystyrene–propionitrile PS–PPN where PPN is a poor solvent we demonstrated11 the continuity of the phase diagram in the region below P = 0. In Fig. 6 we show cloud point data taken for a 0.20 mass fraction PS solution W = 22 000) using the Berthelot technique (2 > P/MPa > (M 20.6) comparing those results with a set of cloud point measurements at higher pressure (5 > P/MPa > 0.1) taken by another technique.26 The two data sets agree for both the upper and the lower consolute branches agreement is nicely within the combined experimental error.The results confirm the idea that the equation of state describing this solution passes smoothly and continuously across the zero pressure isobar and into the region of negative pressure. Thus it is physically reasonable to compare properties of solutions in regions of positive and negative pressure using continuous smoothly varying functions. For example it may be convenient to represent an isopleth (including the critical isopleth) using an expansion about the hypercritical origin even if that origin is found at negative pressure. We have found such expansions a Fig. 6 Polymer–solvent demixing at negative and positive pressures. Cloud points for propionitrile–polystyrene solutions.The solid circles show results obtained by the Berthelot technique for MW = 22,000 Mw/Mn = 1.03 mass. Fraction PS = 0.20. The x values show results obtained at positive pressure using another technique.26 convenient way to represent solution properties even when the hypercritical origin lies so deep as to be experimentally inaccessible or is below the cavitation tension.25 In the second case,12 PS–methyl acetate (PS–MA) we examined the q-solvent–poor solvent transition which is expected to occur at negative pressure (refer to the discussion around Fig 2 and 3). MA is a q-solvent at ordinary pressure and the transition corresponds to a merging of the UCST and LCST at negative pressure. For a solution PS MW = 2 3 106 the point of hypercritical juncture of the upper and lower branches is estimated to lie below 25 MPa and we were unable to reach low enough negative pressures to directly observe the joining point of the two branches (either for this sample or one of MW = 2 3 107 where Phypercr should be somewhat smaller in magnitude).However we were able to demonstrate continuity of the cloud point curve well into the region of negative pressure thus establishing continuity of state and the likely merging of the UCST and LCST branches. hypercr 4 Summary The important idea presented by these negative pressure experiments is that continuity of state extends across the P = 0 boundary into the negative pressure region where solutions are under tension. In this line of thinking the upper and lower UCST and LCST demixing branches share a common cause.The approach therefore forces a certain broadening in the outlook to be employed in describing polymer solution thermodynamics and this has been very useful. One immediate and practical extension was the development of a scaling description of polymer demixing in the (T,X = MW21/2)ycrit P plane which employs an expansion about the hypercritical origin X even for Xhypercr < 0. The approach is in exact analogy of the expansion about Phypercr discussed in this paper in the (T,P)ycrit X>0 plane even for Phypercr < 0. The considerable advantages of this scaling description are discussed at length in a recent review from this laboratory.25 5 Acknowledgement This research was supported by the United States Department of Energy Division of Materials Sciences.6 References 1 M. Marvan Negative Absolute Temperature Iliffe Books Ltd. London 1966. 2 G. S. Kell Am. J. Phys. 1983 51 1038. 3 F. Donny Ann. Chim. Phys. 1846 16 167. (The quotation was translated by G. S. Kell Ref. 2) 4 M. Berthelot Ann. Chim. Phys. 1850 30 232. 5 O. Reynolds Mem. Proc. Manchester Liter. Philos. Soc. 1882 7 1. 6 D. H. Trevena Cavitation and Tension in Liquids Adam Hilger Bristol 1987. 7 A. T. Hayward Am. Scientist 1971 59 434. 8 R. E. Apfel Sci. Am. 1972 227 58. 9 A. Imre and W. A. Van Hook J. Polym. Sci. B 1996 34 751 and references therein. 10 B. A. Wolf and G. Blaum Macromol Chem. 1976 177 1073. 11 A. Imre and W. A. Van Hook J.Polym. Sci. B 1994 32 2283. 12 A. Imre and W. A. Van Hook J. Polym. Sci. B 1997 35 1251. 13 L. Zeman and D. Patterson J. Phys. Chem. 1972 76 1214. 14 S. J. Henderson and R. J. Speedy J. Phys. E 1980 13 778. 15 J. C. Fisher J. Appl. Phys. 1948 19 1062. 16 Although Huygens’ experiment was performed over three hundred years ago maybe it is not correct to speak about a ‘three hundred year’ history. Huygens (and some of his followers like Papen and Boyle) did not realize they were observing negative pressure. That realization only came later (in the 19th century). 17 C. A. Angell and Z. Qing Phys. Rev. B 1989 39 8784. 18 E. Piorkowska and A. Galeski J. Polym. Sci. B 1993 31 1285. 19 G. W. Scherer and D. M. Smith J. Non-Cryst. Solids 1995 189 197. 20 S. J. Henderson and R. J. Speedy J. Phys. Chem. 1987 91 3069. 21 C. Rangel-Nafaile A. N. Metzner and K. F. Wissbrun Macromolecules 1984 17 1187. 22 Y. Ohde M. Ikemizu H. Okamoto T. Yokoyama and S. Shibata J. Phys. D 1988 22 1721. 23 J. L. Green D. J. Durben G. H. Wolf and C. A. Angell Science 1990 249 649. 24 It is particularly appropriate to study polymer–solvent systems using the Berthelot technique because according to Flory (P. J. Flory Principles of Polymer Chemistry Cornell University Press Ithaca NY 1953) it was this scientist who originally coined the term polymerization. ‘Simon reported in 1839 the conversion of styrene to a gelatinous mass. And Berthelot applied the term polymerization to the process in 1866.’ Our experiments were carried out at near-critical concentrations where the cloudpoint spinodal and binodal (equilibrium) curves converge. At off-critical concentrations the cloud point curve lies somewhere in the metastable region i.e. between binodal and spinodal loci. 25 A. Imre and W. A. Van Hook J. Phys. Chem. Ref. Data 1996 25 637. 26 M. Luszczyk and W. A. Van Hook Macromolecules 1996 29 6612. 27 R. S. Vincent Proc. Roy. Soc. 1943 55 41. 28 L. J. Briggs J. Appl. Physics. 1950 21 721. Received 19th June 1997 Accepted 20th October 1997 123 Chemical Society Reviews 1998 volume 27

ISSN:0306-0012    

DOI:10.1039/a827117z

出版商:RSC    

年代:1998

数据来源: RSC

摘要:

Seven-coordinate halocarbonyl complexes of the type [MXY(CO)3(NCMe)2] (M = Mo W; X Y = halide pseudo halide) as highly versatile starting materials Paul K. Baker Department of Chemistry University of Wales Bangor Gwynedd UK LL57 2UW E-mail CHS018@BANGOR.AC.UK ABCDEF}2 dimer The fortuitous discovery of the seven-coordinate complexes [MI2(CO)3(NCMe)2] (M = Mo or W) is described; the wideranging chemistry of these compounds is discussed together with the use of the effective atomic number rule in predicting their reactions. The four structural types of seven-coordinate halocarbonyl complexes derived from [MI2(CO)3(NCMe)2] are described. The applications of the [MXY(CO)3(NCMe)2] (X = Y = Br I; X = Cl Br Y = I; X = Cl Y = GeCl3 SnCl3) system in both alkene metathesis catalysis and nitrogen fixation are also discussed.The synthesis of organometallic complexes containing six different ligands together with the preparation and stereochemical properties of the first structurally characterised {M [Cl(CO)(PPh3)(h2-MeC2Me)Mo(m-NCS)(m-SCN)MoCl- (CO)(PPh3)(h2-MeC2Me)] are described. 1 Introduction The availability of suitable starting materials has always been of considerable importance to the synthetic chemist. Since the first halocarbonyl donor ligand complex of molybdenum(ii) namely [MoI2(CO)2(diars)] {diars = C6H4(AsMe2)2-1,2} was prepared by Nigam and Nyholm in 1957 using reaction (1),1 six- and Paul K. Baker (born Isleworth Middlesex 1954) obtained his BSc (1976) and PhD (1979) from the University of Bristol.His thesis entitled ‘The Synthesis and Reactions of Cationic Organotransition-metal Complexes’ was obtained under the guidance of Neil G. Connelly and Michael Green. He held temporary Lectureships at the University of Strathclyde (1980) and the University of Manchester (1981) followed by a Lectureship in Chemistry at Thames Polytechnic (now the University of Greenwich). In 1984 he was appointed to a Lectureship in Chemistry at the University of Wales Bangor and was promoted to Senior Lecturer in 1989 and subsequently to a Readership in 1994. He is co-author of over 180 research publications mainly in the field of the organometallic chemistry of molybdenum and tungsten. He has given lectures on his research in over 40 UK Universities and Industrial Companies and he has lectured extensively in Australia Austria Belgium France Germany Holland Hong Kong Hungary Italy Poland Portugal South Africa Spain and the USA on various aspects of his research work.He is a Fellow of the Royal Society of Chemistry. In July 1997 the author was awarded the Teaching Fellowship for the Faculty of Science at the University of Wales Bangor. seven-coordinate halocarbonyl complexes have received considerable attention.2 [Mo(CO)4(diars)] + I2?[MoI2(CO)2(diars)] + 2CO (1) In the early days the two most commonly used routes to seven-coordinate halocarbonyl complexes were (i) the oxidation of complexes of molybdenum(0) or tungsten(0) containing strong field ligands and (ii) reactions of the reactive halobridged dimers [{M(m-X)X(CO)4}2] (M = Mo W; X = Cl Br I) with strong field ligands.An early example of type (i) is given in reaction (2).3 [M(CO)4(2,5-dithiahexane)] + X2? [MX2(CO)3(2,5-dithiahexane)] + CO (2) M = Mo W; X = Br I In 1966 Colton and Tomkins4 described the synthesis of the chloro-bridged dimer [{Mo(m-Cl)Cl(CO)4}2] [reaction (3)]. 278 °C 2[Mo(CO)6] + 2Cl2 ——–? [{Mo(m-Cl)Cl(CO)4}2] + 4CO (3) From 1966 to 1971 Colton et al. investigated the chemistry 4}2] (M = Mo W; X = Cl Br I) with of [{M(m-X)X(CO) strong field ligands to give a range of six- and seven-coordinate complexes of molybdenum(ii) and tungsten(ii) and some of their results are published in a review article.5 The molecular structure of the tungsten bromo-bridged dimer [{W(m- Br)Br(CO)4}2] has been crystallographically determined by Cotton et al.,6 and shows each tungsten atom to be in a capped octahedral environment with a carbonyl group capping an octahedral face (Fig.1). In 1972 Westland and Muriithi7 described the reactions of the halo-bridged dimers [{Mo(m- X)X(CO)4}2] (X = Cl Br) with weaker field ligands such as pyridine tetrahydrofuran and acetonitrile which eventually gave the non-carbonyl containing products [MoX3L3] and zerovalent [Mo(CO)6] via disproportionation of the six-coordinate molybdenum(ii) intermediates [MoX2(CO)2L2]. They were apparently unable to isolate these intermediates in a pure state. In 1962 Tate Knipple and Augl8 published the seminal paper on the synthesis of the important starting materials fac- [M(CO)3(NCMe)3] (M = Cr Mo or W) by refluxing [M(CO)6] Chemical Society Reviews 1998 volume 27 Fig 1.Molecular structure of [{W(m-Br)Br(CO)4}2]6 125 in acetonitrile for various periods of time. These zero-valent tris(acetonitrile) complexes have been used as starting materials for the preparation of a wide range of organotransition-metal complexes including p-allyl halocarbonyl complexes.2 In their paper,8 Tate Knipple and Augl showed that the reaction of [W(CO)3(NCMe)3] with iodine in methanol evolved three moles of gas and gave a non-carbonyl containing product. They did not isolate the seven-coordinate bis(acetonitrile) complex [WI2(CO)3(NCMe)2]. In view of these observations we decided to reinvestigate the halogen oxidation of fac- [M(CO)3(NCMe)3] (M = Mo or W) by reacting them in acetonitrile rather than methanol in the hope of obtaining the seven-coordinate complexes [MX2(CO)3(NCMe)2] with two 3(NCMe)2] were prepared.9 More recently,10 we have potentially labile acetonitrile ligands.These reactions (4) first carried out in November 1984 have been extremely successful and quantitative yields of the diiodo-complexes [MI2- (CO) developed a better route to the dibromo-complexes [MBr2(CO)3(NCMe)2] by carrying out the reaction of fac- [M(CO)3(NCMe)3] (prepared in situ) with Br2 at 278 °C. 0 °C fac-[M(CO)3(NCMe)] + X2 —–? [MX2(CO)3(NCMe)2] + NCMe (4) M = Mo or W; X = Br or I We have also studied the oxidation of fac-[Mo- (CO)3(NCMe)3] (prepared in situ8) with a range of other oxidizing agents XY (XY = ICl,11 IBr,12 GeCl4,13 SnCl4 14) to give [MoXY(CO)3(NCMe)2] (X = Cl Y = I; X = Br Y = I; X = Cl Y = GeCl3; X = Cl Y = SnCl3).During the past 12 years we have prepared and characterised over 1900 new organometallic complexes derived from the highly versatile starting materials [MXY(CO)3(NCMe)2].2 This article will describe the versatility of the [MI2- (CO)3(NCMe)2] complexes in Section 2 and discusses how the effective atomic number rule can be used to predict the products of these reactions. Section 3 will give an overview on the structures of seven-coordinate complexes derived from [MI- 2(CO)3(NCMe)2] and Section 4 discusses the applications of the [MXY(CO)3(NCMe)2] system in alkene metathesis and nitrogen fixation research.Section 5 discusses the synthesis of complexes of molybdenum(ii) containing six-different monodentate ligands. The conclusion (Section 6) discusses current and future developments of the [MXY(CO)3(NCMe)2] system. 2 Investigation of the chemistry of the highly versatile complexes [MI2(CO)3(NCMe)2] Although the chemistry of the [MXY(CO)3(NCMe)2] (M = Mo or W; X = Y = Br I; for M = Mo only X = Cl Y = I; X = Br Y = I; X = Cl Y = GeCl3 SnCl3) has been extensively investigated,2 and differences in reactivity are apparent as discussed in Section 5 this section will concentrate on the diiodo-complexes [MI2(CO)3(NCMe)2] which have received the most attention. The complexes [MI2(CO)3(NCMe)2] (M = Mo or W) have proved ideal starting materials for a wide range of reactions.Although they are air-sensitive in both the solid state and solution they can be weighed in air and standard vacuum/Schlenk line techniques are very suitable for carrying out all manipulations with this system. These starting materials can be easily prepared in multigram quantities in a pure state. We have modified Tate Knipple and Augl’s8 early report for the synthesis of fac-[M(CO)3(NCMe)3] complexes by reacting [Mo(CO)6] under reflux in NCMe for 24 h and [W(CO)6] under reflux in NCMe for 72 h. It is essential to have very careful and thorough degassing of the [M(CO)6]–NCMe mixture before refluxing as shown on the BBC Open University video made in Bangor in 1988,15 for the Third Level Inorganic Chemistry Course CHEM 777 S343.This video together with an associated student book15 uses the [MXY(CO)3(NCMe)2] system to illustrate Schlenk line techniques spectroscopic and Chemical Society Reviews 1998 volume 27 126 X-ray techniques and various other techniques used in modern day synthetic inorganic chemistry. Scheme 1 illustrates the diverse chemistry of [MI2(CO)3- (NCMe)2]. Although Scheme 1 is reasonably complete there are many subtle reactions which are not given in the scheme. These reactions depending on the electronic and steric effects of the ligands added can give a diverse range of different products. The top half of the scheme from [MI2(CO)3(NCMe)2] describes the reactions with neutral and anionic ligands whereas the bottom half concentrates on the alkyne derivatives of [MI2(CO)3(NCMe)2].We have found that the complexes [MI2(CO)3(NCMe)2] react with almost any potential ligand in the current chemical catalogues including poor donor ligands for example BiPh3.16 The starting materials [MI2- (CO)3(NCMe)2] and all the reaction products shown in Scheme 1 obey the effective atomic number rule. Although over 1900 new complexes have been made from the [MXY(CO)3(NCMe)2] system which has provided a good training for research students in the synthesis and characterisation of moderately air-sensitive complexes; the main significance is in the catalytic activity of many of these complexes (see Sections 4 and 6). Alkyne ligands can donate two or four electrons to a transition-metal centre,17 and in the early transition-metal complexes alkynes generally utilize both their filled pporbitals and donate four electrons to the metal.It is interesting to note that in the bis(alkyne) complexes [WI2(CO)(NCMe)( h2-RC2RA)2]18 shown in Scheme 1 each alkyne must donate an average of three electrons to the metal for the complexes to obey the effective atomic number rule. The 13C NMR chemical shifts for the alkyne carbon atoms in the complexes [WI2(CO)(NCMe)(h2-RC2RA)2] are generally in the region of 150 ppm which agrees with Templeton and Ward’s17 correlation indicating that the alkynes are donating an average of three electrons to the metal in these systems. The reaction products of [WI2(CO)(NCMe)(h2-RC2RA)2] with both phosphine19 and phosphite20 donor ligands give mono(alkyne) products [WI2(CO)L2(h2-RC2RA)] which have alkyne carbon chemical shifts above 200 ppm which indicates17 they are donating four electrons to the metal thus enabling them to obey the effective atomic number rule.Very recently,21 we have described the reactions of [WI2(CO)3L2] {L = PEt3 PPh3; L2 = Ph2P(CH2)2PPh2} with Na2[S÷S] {S÷S = C3S5 (C3S5 = 4,5-disulfanyl-1,3-dithiole-2-thionate) or maleonitriledithiolate} or H2bdt (bdt = benzene-1,2-dithiol) in acetonitrile and ethanol to give the six-coordinate complexes [W(S÷S)(CO)2L2] which do not obey the effective atomic number rule. Hence although the vast majority of our reactions are predicted by the effective atomic number rule there are several exceptions to the rule.Research workers in other countries have also used the [MI2(CO)3(NCMe)2] complexes as starting materials in their research. For example Krishnamurthy et al.22 have reported the reactions of the diphosphazane ligands RN{P(OPh)2}2 (R = Me Ph) with [MI2(CO)3(NCMe)2] to afford the acetonitrile replaced seven-coordinate complexes [MI2(CO)3- ({P(OPh)2}2NR)] which has been crystallographically characterised for M = W R = Ph. Very recently Cano et al.23 have described the reactions of [MoI2(CO)3(NCMe)2] with the hydrotris(3,5-dimethylpyrazol-1-yl)borate ligand to yield a range of products depending upon the solvent and the initial concentration of the starting materials including [Mo4O4(m3- O)2(m2-O)2(m2-OH)2(HpzMe2)6]I2·4NCMe which has been crystallographically characterised.3 Structures of seven-coordinate complexes derived from [MI2(CO)3(NCMe)2] Seven-coordination is an unusual geometry in transition-metal chemistry and there are four main types of structure exhibited by seven-coordinate complexes; these are (i) capped octahedral, Scheme 1 M = Mo or W; for the reaction of [MI2(CO)3(NCMe)2] with donor ligands described in the top half of the page L LA and LB represent a wide range of neutral carbon nitrogen phosphorus arsenic antimony bismuth oxygen and sulfur donor ligands. L÷L represents a neutral bidentate phosphine donor ligand N÷N represents a neutral bidentate nitrogen donor ligand such as 2,2A-bipy 1,10-phen etc.For the top and bottom half of the scheme S2CX2 (X = NR2 R = Me Et CH2Ph etc.; X = OEt SEt etc.). For the reactions of the bis(alkyne) complexes [WI2(CO)(NCMe)(h2-RC2RA)2] (R RA = Me Ph etc.) L L÷L and L÷L÷L represent mono- bi- and tri-dentate phosphine ligands respectively and X2; X = Cl Br I NO2 NO3 NCS OH SR (R = Et But Ph CH2Ph). 2- (ii) capped trigonal prismatic (iii) the so-called ‘4 3 geometry’ and (iv) pentagonal bipyramidal geometry. Several important reviews concerned with the structures of seven-coordinate complexes have been published,2,24,25 which also discuss the structures of the [MX2(CO)3L2] type complexes. In reality some of the structures we have found derived from the [MI2(CO)3(NCMe)2] system do not exhibit one of the four perfect geometries described above but crystallise in some intermediate structure often in between two structural types.It has been shown that very few coordination polyhedra have regular structures and the majority lie on the reaction coordinates connecting the ideals. The majority of the structures which have been determined of seven-coordinate complexes derived from [MI2(CO)3(NCMe)2] are capped octahedral or distorted capped octahedral although all four principal geometries have been observed. In order to clarify the main geometries of seven-coordinate complexes derived from [MI (CO)3(NCMe)2] accurate coloured diagrams of crystal structures together with simple model pictures of complexes for all four representative structures are shown in Figs.2–5. Fig. 2 (a) Molecular structure of [WI2(CO)3(NCMe)2].26 (b) Simple representation of the capped octahedral structure of [WI2(CO)3(NCMe)2]. The molecular structure of the tungsten starting material [WI2(CO)3(NCMe)2] has been crystallographically determined 127 Chemical Society Reviews 1998 volume 27 Fig. 3 (a) Molecular structure of [WI2(CO)3(NCMe)(SbPh3)].10 (b) Simple representation of the capped trigonal prismatic geometry of [WI2- (CO)3(NCMe)(SbPh3)]. Fig. 4 (a) Molecular structure of the cation [WI(CO)3([9]aneS3)]+.28 (b) Simple representation of the 4 3 geometry or piano-stool geometry of [WI(CO)3([9]aneS3)]+. (Fig. 2),26 and has distorted capped octahedral geometry with a carbonyl ligand in the unique capping position which is capping an octahedral face containing two carbonyl ligands and one of the iodine atoms as shown in Fig.2(b). The large iodide ligands are trans to each other in this complex. Colton and Fig. 5 (a) Molecular structure of [WI(acac)(CO)2(PEt3)2].29 (b) Simple representation of the pentagonal bipyramidal geometry of [WI(acac)(CO)2- (PEt3)2]. Chemical Society Reviews 1998 volume 27 128 Kevekordes27 have shown how the solution state 13C NMR spectra of capped octahedral complexes containing a capping carbonyl ligand can be correlated with the crystallographically determined solid-state structure. For example the low temperature 13C NMR spectrum (270 °C 250 MHz CD2Cl2) of [WI2(CO)3(NCMe)2],26 has two carbonyl resonances at d = 228.48 and 202.36 with an intensity ratio of 1 2.The lower field resonance is due to the capping carbonyl group.27 The resonance at d = 202.36 is in the typical range for octahedral carbonyl ligands. The room temperature 13C NMR spectrum (+25 °C 250 MHz CD2Cl2) of [WI2(CO)3(NCMe)2] has a single carbonyl resonance which as expected suggests the complex is fluxional due to the small amount of energy required for these seven-coordinate complexes to rearrange. Many seven-coordinate halocarbonyl complexes exhibit this fluxional behaviour.2,27 The other three less commonly observed geometries of complexes derived from [MI2(CO)3(NCMe)2] are exemplified by (a) [WI2(CO)3(NCMe)(SbPh3)],10 which has a distorted capped trigonal prismatic geometry with one triangular face having three CO groups and the other having I(2) Sb and N with I(1) capping the rectangular face defined by C(1) I(2) N and C(3) {see Fig.3(a) and (b)}; (b) the S3-bonded macrocyclic complex [WI(CO)3([9]aneS3)][BPh4]28 which can be considered to have the so called ‘4 3 geometry’ with the three sulfur atoms forming the seat and the three carbonyl ligands and an iodide group forming the four legs of this often called ‘piano stool’ geometry {refer to Fig. 4(a) and (b)}; and (c) the least commonly observed structure in this system which is the pentagonal bipyramidal geometry and which has only been found in very recently29 observed structures of [MI(acac)- (CO)2(PEt3)2] (M = Mo W) {for M = W see Fig. 5(a) and (b)}. It should be noted that many seven-coordinate lanthanide and actinide complexes do exhibit the pentagonal bipyramidal geometry.24 There is obviously a complex interplay between steric and electronic effects and multidenticity of the ligands attached to the metal in these systems which governs the geometry of these complexes.It is important to note that the geometry observed in the solid state appears to be preserved in solution as evidenced by 13C NMR spectroscopy.27 4 Applications of the complexes [MXY(CO)3(NCMe)2] and their derivatives Throughout the course of our work with the [MXY- (CO)3(NCMe)2] system we have been conscious of the importance of the potential applications of these complexes in other areas of research. Two of the main potential applications of the complexes [MXY(CO)3(NCMe)2] and their derivatives we have been exploring are in (i) alkene metathesis catalysis and (ii) nitrogen fixation research.Our interest in developing the chemistry of [MXY- (CO)3(NCMe)2] with a very wide range of ligands results partly from a paper published by Bencze and Kraut-Vass30 in 1985 which describes the catalytic activity of the halocarbonyl complexes [MX2(CO)3L2] (M = Mo or W; X = Cl or Br; L = PPh3 or AsPh3). Complexes of this type were found to be single component catalysts for the ring opening polymerisation of norbornene and norbornadiene.30 The best catalyst was found to be [WCl2(CO)3(AsPh3)2] which polymerises norbornadiene to give a vinylcyclopentene polymer with a high cis-content. Bencze and Kraut-Vass also found30 that the mechanism of these reactions (5) involved loss of L in the rate-determining step followed by coordination of the alkene (6).RDS (5) [MX2(CO)3L2] [| [MX2(CO)3L] + L [MX2(CO)3L] + alkene [| [MX2(CO)3L(h2-alkene)] (6) Further studies by Bencze et al.31 showed that the next step in the mechanism involves a 1,2-hydrogen shift on the coordinated alkene to give a carbene intermediate. In view of these observations we have been studying the catalytic activity of [MXY(CO)3(NCMe)2] and many of their derivatives shown in Scheme 1 towards the metathesis of a wide range of acyclic and cyclic alkenes with considerable success. Since Bencze and Kraut-Vass30 showed the rate-determining step in these catalytic reactions involved loss of donor ligand {see reaction (5)} it is significant that in our system the complexes [MXY- (CO)3(NCMe)2] and [MXY(CO)3(NCMe)L] have at least one labile acetonitrile ligand and many of the complexes in Scheme 1 including the alkyne complexes are highly active alkene metathesis catalysts.32 For example the complexes [MBr2(CO)3(NCMe)2] (M = Mo or W) and [WBr2(CO)3(NCMe)( EPh3)] (E = P As Sb) all rapidly initiate the room temperature polymerisation of norbornadiene.32 Since the nature of L can be varied inexhaustibly in this system the stereochemistry of the products can in principle be controlled in these reactions.The chemical industry puts a high emphasis on water soluble homogeneous catalysts and during the course of our work in conjunction with Dr Aidan J.Lavery ZENECA Specialties we have discovered the cheap and readily available water solubilising substituted pyridine ligands such as sodium [pyridine- 4-carboxylate] which forms completely water-soluble complexes such as [WI2(CO)3(4-NaO2CC5H4N)2].33 This and related water-soluble seven-coordinate complexes are being explored as potential biphasic catalysts. Over the years much research has been directed to synthesising simple models that mimic the active site of the nitrogenase enzyme which contains molybdenum iron and sulfur in the active site. Recently the X-ray structure analysis of the three forms of the active site of the FeMo protein which was isolated from Azobacter vinelandii and Clostridium pasteurianum has been described,34 and is shown in Fig.6. From our point of view CH2CO2 – S S S Fe Fe O C CH2CH2CO2 – Mo CysFe O C S S Fe S Fe O His Fe Fe S 2- S ABCDEF] complexes [MoCl(GeCl3)- S Fig. 6 Structure of the active site of FeMoco from an X-ray structural analysis34 it has a molybdenum centre with three attached sulfur atoms together with a bidentate anionic oxygen donor ligand and a histidine group. In conjunction with Dr Raymond L. Richards et al.28,29 at the Nitrogen Fixation Laboratory we have been investigating the [MXY(CO)3(NCMe)2] system in nitrogen fixation research by firstly investigating the chemistry of [MXY(CO)3(NCMe)2] and derivatives with a wide variety of neutral and anionic oxygen and sulfur donor ligands and with iron–sulfur clusters such as [Fe4Cp4S6] in order to attempt to mimic the active site of nitrogenase shown in Fig.6. In order to get three sulfur atoms attached to molybdenum or tungsten we have used28 the reactions of equimolar amounts of [MI (CO)3(NCMe)2] with for example the macrocycle [9]aneS3 in the presence of Na[BPh4] to give the cationic complexes [MI(CO)3([9]aneS3)][BPh4]. This has been crystallographically characterised for M = W shown in Fig. 4(a). We are also investigating the reduction of the sulfur donor ligand complexes in the presence of dinitrogen in order to form dinitrogen complexes containing sulfur donor ligands. 5 Complexes containing six different monodentate ligands derived from [MXY(CO)3(NCMe)2] Very few octahedral transition-metal complexes containing six different monodentate ligands have been described.An early example of a [MABCDEF] complex is [PtCl(Br)I- (NO2)(NH3)(py)] reported in 1958.35 In order to illustrate the subtle differences in the chemistry of [MI2(CO)3(NCMe)2] and other [MXY(CO)3(NCMe)2] type complexes this section describes the strategic synthesis of [MABCDEF] molybdenum complexes and the X-ray crystal structure of a dinuclear complex which has two identical MABCDEF centres. The complete synthetic scheme for the synthesis of the [MABCDEF] complexes [MoCl(GeCl3)(CO)(NCMe)(PPh3)(h2- RC2R)] (R = Me or Ph) is shown in Scheme 2.36 It is interesting to note that reaction of [MoXY(CO)3(NCMe)2] with an equimolar amount of PPh3 in NCMe for X = Y = I gives the acetonitrile displaced product [MoI2(CO)3(NCMe)(PPh3)] whereas for X = Cl Y = GeCl3 the carbon monoxide displaced product [MoCl(GeCl3)(CO)2(NCMe)2(PPh3)] is obtained.This is likely to be due to the combined higher electronegativity of the Cl and GeCl3 groups which weakens the M–C bond due to decreased back-bonding and strengthens the Mo–N bond due to increased s-bonding compared with the two iodo ligands. As previously discussed,17 alkynes can act as ‘four-electron donors’ and hence the reaction of [MoCl- (GeCl3)(CO)2(NCMe)2(PPh3)] with one equivalent of RC2R yielded the chiral [M Chemical Society Reviews 1998 volume 27 129 Scheme 2 (CO)(NCMe)(PPh3)(h2-RC2R)] via displacement of a CO and an NCMe ligand. There are 15 possible simple geometrical isomers of an [MABCDEF] octahedral complex and each one will have its own diastereoisomer.Over the years many hitherto unsuccessful attempts have been made to grow suitable single crystals for X-ray analysis of [M 2Me)] with an equimolar amount of [NBun 4][NCS] ABCDEF] complexes derived from [MoCl(GeCl3)(CO)(NCMe)(PPh3)(h2-RC2R)]. However treatment of [MoCl(GeCl3)(CO)(NCMe)(PPh3)(h2- MeC gave the anionic complex [NBun (CO)(PPh 4][MoCl(GeCl3)(NCS)- 3)(h2-MeC2Me)] which reacted in acetone with a slight trace of water to afford the dimeric {MABCDEF}2 complex 3)(h2-MeC2Me)Mo(m-NCS)(m-SCN)MoCl(CO)- 3)(h2-MeC2Me)]. This is the first {MABCDEF}2 complex to [Cl(CO)(PPh (PPh be structurally characterised (Fig. 7).37 Each molybdenum atom has terminal chloride carbon monoxide triphenylphosphine and but-2-yne ligands and a nitrogen and sulfur of two bridged thiocyanate groups.One of the reasons for preparing mononuclear MABCDEF complexes was to synthesise and characterise new chiral catalysts. However the dinuclear complex above is very interesting from a stereochemical point of view which was not discussed in the original communication.37 The dimer (Fig. 7) is not a chiral molecule as there is a centre of symmetry located at the centre of the eight-membered ring and hence the dimer is the achiral meso form of the molecule (point group Ci). The two Mo atoms are both stereocentres but have opposite absolute configurations which can be classified as C and A using the Cahn–Ingold–Prelog priority rules by analogy with the well known R/S nomenclature system for tetrahedral stereocentres.6 Conclusions and future developments of the [MXY(CO)3(NCMe)2] system The main purpose of this article has been to illustrate how the easily prepared seven-coordinate complexes [MI2- (CO)3(NCMe)2] have proved to be highly versatile starting materials for a wide range of reaction types as shown in Scheme 1. The diversity of this research is also highlighted by the use of a number of other oxidizing agents in place of I2 such as XY (XY = ICl,11 IBr,12 GeCl4 13 or SnCl4 14) which when reacted with fac-[M(CO)3(NCMe)3] (prepared in situ8) give the products [MXY(CO)3(NCMe)2]. The [MXY(CO)3(NCMe)2] Chemical Society Reviews 1998 volume 27 130 Fig.7 Molecular structure of [Cl(CO)(PPh3)(h2-MeC2Me)Mo(m-NCS)(m- SCN)MoCl(CO)(PPh3)(h2-MeC2Me)]37 series of complexes have proved to be extremely useful as starting materials for a wide range of chemistry. I believe the most significant current and future potential uses 3(NCMe)2] system are in homogeneous 3(NCMe)2] and derivatives in alkene metathesis 2(CO)- 2- of the [MXY(CO) catalysis. Apart from the previously described,32 uses of [MXY(CO) (Section 4) which we are continuing to develop; a very recent result38 has shown that the bis(phosphite) complex [MoI {P(OPri)3}2(h2-MeC2Ph)] catalyses the trimerisation of 1-phenylpropyne to give 1,2,4-trimethyl-3,5,6-triphenylbenzene. We are currently exploring the applications of [MoI (CO){P(OPri)3}2(h2-MeC2Ph)] and related complexes for the trimerisation of other alkynes.The uses of the organometallic phosphines such as [WI2(CO){PhP(CH2CH2PPh2)2-P,PA}(h2- MeC2Me)]39 which exists as two diastereoisomers in a single crystal as shown in {Fig. 8(a) and (b)} for the synthesis and applications of mixed-metal catalysts are also being investigated. 7 Acknowledgements I would like to thank all my co-workers who have been involved with the [MXY(CO)3(NCMe)2] system for their skilful experimental work and intellectual contributions. I would also like to thank in particular Dr Michael G. B. Drew University of Reading for his outstanding collaboration over many years and for producing the colour diagrams used in this review. I also thank Professor Michael B.Hursthouse and his collaborators on the EPSRC Crystallographic Service (University of Wales Cardiff) and Dr David L. Hughes (Nitrogen Fixation Laboratory Norwich) for their crystallographic work; Dr Raymond L. Richards (Nitrogen Fixation Laboratory Norwich) and Dr Aidan J. Lavery (ZENECA Specialties) for their valuable collaboration and support of our work over the years. I would like to thank the EPSRC AFRC Royal Society British Council Institute of Terrestrial Ecology and ZENECA Specialties for supporting our research work in this field over the past twelve years. Finally I would like to thank Barbara Kinsella and Fig. 8 Molecular structure of the two diastereoisomers of [WI (CO){PhP(CH2CH2PPh2)2-P,PA}(h2-MeC2Me)]39 2- Caroline Naylor for their cheerful and efficient preparation of this article.8 References 1 H. L. Nigam and R. S. Nyholm Proc. Chem. Soc. 1957 321. 2 P. K. Baker Adv. Organomet. Chem. 1996 40 45 and references therein. 3 H. C. E. Mannerskantz and G. Wilkinson J. Chem. Soc. 1962 4454. 4 R. Colton and I. B. Tomkins Aust. J. Chem. 1966 19 1143. 5 R. Colton Coord. Chem. Rev. 1971 6 269 and references therein. 6 F. A. Cotton L. R. Falvello and J. H. Meadows Inorg. Chem. 1985 24 514. 7 A. D. Westland and N. Muriithi Inorg. Chem. 1972 11 2971. 8 D. P. Tate W. R. Knipple and J. M. Augl Inorg. Chem. 1962 1 433. 9 P. K. Baker S. G. Fraser and E. M. Keys J. Organomet. Chem. 1986 309 319. 10 P. K. Baker M. B. Hursthouse A. I. Karaulov A.J. Lavery K. M. A. Malik D. J. Muldoon and A. Shawcross J. Chem. Soc. Dalton Trans. 1994 3493. 11 P. K. Baker T. Birkbeck S. Br�ase A. Bury and H. M. Naylor Transition-Met. Chem. 1992 17 401. 12 P. K. Baker K. R. Flower H. M. Naylor and K. Voigt Polyhedron 1993 12 357. 13 P. K. Baker and D. ap Kendrick J. Organomet. Chem. 1994 466 139. 14 P. K. Baker and A. Bury J. Organomet. Chem. 1989 359 189. 15 Open University S343 Laboratory Techniques of Inorganic Chemistry 1990 4. 16 P. K. Baker S. G. Fraser and T. M. Matthews Inorg. Chim. Acta 1988 150 217. 17 J. L. Templeton and B. C. Ward J. Am. Chem. Soc. 1980 102 3288. 18 E. M. Armstrong P. K. Baker and M. G. B. Drew Organometallics 1988 7 319. 19 E. M. Armstrong P. K. Baker M. E.Harman and M. B. Hursthouse J. Chem. Soc. Dalton Trans. 1989 295. 20 P. K. Baker E. M. Armstrong and M. G. B. Drew Inorg. Chem. 1989 28 2406. 21 P. K. Baker M. G. B. Drew E. E. Parker N. Robertson and A. E. Underhill J. Chem. Soc. Dalton Trans. 1997 1429. 22 M. S. Balakrishna S. S. Krishnamurthy and H. Manohar Organometallics 1991 10 2522. 23 M. Cano J. A. Campo J. V. Heras E. Pinilla and A. Monge Polyhedron 1996 15 1705. 24 M. G. B. Drew Prog. Inorg. Chem. 1977 23 67 and references therein. 25 M. Meln�ýk and P. Sharrock Coord. Chem. Rev. 1985 65 49 and references therein. 26 M. G. B. Drew P. K. Baker E. M. Armstrong S. G. Fraser D. J. Muldoon A. J. Lavery and A. Shawcross Polyhedron 1995 14 617. 27 R. Colton and J. Kevekordes Aust. J. Chem. 1982 35 895 and references therein. 28 P. K. Baker S. J. Coles M. C. Durrant S. D. Harris D. L. Hughes M. B. Hursthouse and R. L. Richards J. Chem. Soc. Dalton Trans. 1996 4003. 29 P. K. Baker A. I. Clark M. G. B. Drew M. C. Durrant and R. L. Richards J. Organomet. Chem. 1997 549 193. 30 L. Bencze and A. Kraut-Vass J. Mol. Catal. 1985 28 369. 31 L. Bencze A. Kraut-Vass and L. Pr�okai J. Chem. Soc. Chem. Commun. 1985 911. 32 D. J. Muldoon PhD Thesis University of Wales Bangor 1994. 33 P. K. Baker A. E. Jenkins A. J. Lavery D. J. Muldoon and A. Shawcross J. Chem. Soc. Dalton Trans. 1995 1525. 34 D. Sellmann Angew. Chem. Int. Ed. Engl. 1993 32 64 and references therein. 35 L. N. Essen F. A. Zakharova and A. D. Gel’man Zh. Neorg. Khim. 1958 3 2654. 36 P. K. Baker and D. ap Kendrick J. Chem. Soc. Dalton Trans. 1993 1039. 37 P. K. Baker M. E. Harman D. ap Kendrick and M. B. Hursthouse Inorg. Chem. 1993 32 3395. 38 P. K. Baker M. G. B. Drew and D. S. Evans unpublished results. 39 P. K. Baker S. J. Coles D. E. Hibbs M. M. Meehan and M. B. Hursthouse J. Chem. Soc. Dalton Trans. 1996 3995. Received 28th July 1997 Accepted 15th October 1997 131 Chemical Society Reviews 1lume 27

ISSN:0306-0012    

DOI:10.1039/a827125z

出版商:RSC    

年代:1998

数据来源: RSC

摘要:

NMR studies of carbohydrate–protein interactions in solution A. Povedaa and J. Jim�enez-Barberob a onoma de Madrid Cantoblanco-28047 Madrid Spain RMN-SIdI Universidad Aut� b anica CSIC Juan de la Cierva 3 28006 Madrid Spain Instituto Qu�ýmica Org� This review provides an overview of the current methods of NMR spectroscopy that may be used to obtain information about the conformational aspects forces and structural motifs that govern the interactions between proteins and carbohydrates in solution. 1 Introduction Carbohydrates differ from the other classes of biomolecules in that their constituting moieties (monosaccharides) may be connected to one another by a great variety of linkage types. In addition they can be highly branched thus allowing oligosaccharides to provide an almost infinite array of structural variations.The decoding process of the existing information in oligosaccharide structures involves their recognition by other biomolecules. Thus they are most often specifically recognized by proteins (so called lectins) and these interactions may mediate a particular biological response such as host–parasite interactions fertilization autoimmune disorders and cellular differentiation.1 Therefore the study of how oligosaccharides are recognised by the binding sites of lectins enzymes and antibodies is a topic of major interest. It is evident that knowledge of the three dimensional structure of these biomolecules (proteins and carbohydrates) could assist in the design of new carbohydrate-based therapeutic agents.Current technical facilities and biophysical techniques mainly X-ray crystallography have allowed access to detailed information on the three dimensional structure of protein–carbohydrate complexes. 2 These data complemented mainly by those obtained through titration microcalorimetry have allowed postulates on the major factors involved in these interactions to be made.3 Hydrogen bonds and van der Waals forces often including packing of a hydrophobic sugar face against aromatic amino acid side chains are the usual factors invoked (Scheme 1). The relative importance of each type of force depends on the particular protein and this issue remains a topic of discussion. Jes�us Jim�enez-Barbero (Madrid 1960) studied Chemistry at University Autonoma Madrid.He received his PhD with M. Bernab�e (1987). He has performed postdoctoral studies at the CERMAV-CNRS Grenoble; University of Zuerich; National Institute for Medical Research London; and Carnegie Mellon University Pittsburgh. He is Senior Research Scientist of the National Research Council of Spain (CSIC). He has published about 100 papers on conformational studies of free and protein-bound carbohydrates by using NMR and molecular mechanics. His Jes�us Jim�enez-Barbero Scheme 1 Schematic view of van der Waals and hydrogen bond interactions between amino acid residues of a protein and a monosaccharide On the basis of crystal structures of a variety of complexes the amino acids most commonly involved in hydrogen bonds with carbohydrates are known to be Asp Asn > Glu > Arg His Trp Lys > Tyr Gln > Ser Thr.On the other hand those most usually observed in van der Waals interactions are Trp Phe Tyr Leu Val and Ala; that is those with aromatic or aliphatic side chains (Scheme 2). The ability to bind any one type of sugar has evolved independently in diverse lectin frameworks. In turn families of lectins that share common structural features often contain members that recognise different groups of sugars. Although lectins bind monosaccharides rather weakly these proteins employ common strategies for enhancing both the specificity and the affinity of their interactions for more complex oligosaccharide ligands. Thus by using different lectin subsites and/or subunits it is possible to uncover these enhancements (Scheme 3) as several contacts between a given lectin and several carbohydrates or vice versa may take place.In addition a dramatically increased affinity for oligosaccharides may result from clustering of simple binding research interest are structural studies of biomolecules and molecular recognition processes by using NMR and molecular modeling protocols. Ana Poveda (Madrid 1965) studied Chemistry at University Autonoma Madrid. She received her PhD with J. Jim�enez- Barbero on NMR studies of oligosaccharide dynamics. She is in charge of the NMR facility of the University. Her research interests are in applied and methodological NMR. Ana Poveda 133 Chemical Society Reviews 1998 volume 27 Scheme 2 Lateral chains of the different amino acid residues more frequently encountered in the binding sites of protein–carbohydrate complexes.(a) Amino acid residues involved in hydrogen bond interactions. (b) Amino acid residues involved in van der Waals interactions. sites in oligomers of lectins. Apart from the direct interactions between atoms of both the protein and carbohydrate other molecules are often involved in the formation of the complex. Water molecules either located in the binding site or at the surface of the protein have also been shown to provide additional interactions which help to stabilize the complex and to achieve higher selectivity.2 In addition different ions have also been shown to be of crucial importance for the establishment of protein–carbohydrate molecular complexes.Calcium and manganese are encountered in the legume lectins while the animal C-lectins are also calcium-dependent.2,4 Until very recently the large size of carbohydrate-binding proteins (lectins antibodies and enzymes) has precluded direct study using NMR. However in the last few years researchers have begun to apply NMR to the study of the molecular recognition processes involved at different levels of complexity. Although it has already been mentioned that X-ray crystallography has achieved many results in this field oligosaccharides often present on glycoproteins or on protein–carbohydrate complexes are less amenable to crystallography due to their intrinsic flexibility.Moreover carbohydrates also exhibit greater dynamic fluctuations than proteins and therefore NMR measurements may well offer new insights into the conformation of bound oligosaccharides and/or the corresponding dynamic timescales. Chemical Society Reviews 1998 volume 27 134 Thus using NMR (Table 1) it is possible to deduce the specificity and affinity of binding association constants and equilibrium thermodynamic parameters (titration experiments). It may also be possible to deduce which amino acid residues are involved in binding (titration NOESY CIDNP experiments) even the three dimensional structure of the bound carbohydrate and/or the protein (NOESY experiments). From the dynamic point of view it could also be possible to say something about the ligand exchange timescale and the size of the complex (relaxation measurements).Table 1 NMR protocols to study protein–carbohydrate complexes. The information provided from these studies is also shown ] , Titration CIDNP TR-NOE Relaxation 3D Structure ] , , ] Scheme 3 Schematic view of the different ways in which carbohydrate binding proteins enhance the affinity towards their ligands. (a) Existence of different subsites. (b) Different domains of a given lectin may be involved in the binding of the same sugar. (c) Different domains of a given lectin may be involved in the binding of different sugars. (d) Different lectin molecules may bind different regions of a multivalent oligosaccharide. (e) Clustering of lectins and oligosaccharides.Many questions concerning protein–carbohydrate interactions are therefore associated with conformational behavior. In many cases not all of the required information can be obtained directly by experimental studies so theoretical molecular modeling is usually required to supplement experimental data both in solution and in the solid state. Different modeling protocols have recently been proposed to locate protein binding sites. Soed to calculate the interaction energies between ligand and receptor. Other methods are designed to explore systematically all the positions and orientations that the sugar may adopt within the binding site.5 Whenever the experimental 3D structure of a protein– carbohydrate complex is known this information may be employed to derive the structures of other complexes which present structural homology.This so-called knowledge-based model building approach has been used with success for several legume lectins.6 From the point of view of the force fields used in conformational analysis of carbohydrates and in protein– carbohydrate interactions there is no general force field at present although a variety of them have been shown to provide satisfactory agreement between experimental and modeled data.7 Force fields specifically developed for sugars are HSEA and PFOS. General molecular mechanics programs adapted for carbohydrates are MM3 CHARMM AMBER GROMOS and TRIPOS.5–7 A general force field which has also been shown to be useful for deriving three dimensional structures of free carbohydrates is CVFF.In several cases different modifications of a given force field are present in the literature.7 2 NMR investigations of protein–carbohydrate complexes Different examples of the study of the structural events that mediate the molecular recognition processes between proteins and carbohydrates have recently been presented using NMR spectroscopy to examine interactions involving lectin- antibody- and enzyme-type receptors. Specific comparisons of the differential binding of natural and modified analogues have also been reported. In some cases protein-induced conformational changes in the oligosaccharide ligand have been observed (Scheme 4). Nevertheless in other cases the lectin selects one of the conformers present in solution or a structure close to the major one existing in solution.Finally there are also cases in which the ligand retains at least part of the flexibility it has in the isolated state. Scheme 4 (a) Schematic view of the glycosidic torsion angles which define the three dimensional shape of an oligosaccharide. (b) Schematic view of putative conformational changes around the f glycosidic torsion angle of a disaccharide. In principle different types of information may be deduced for protein–carbohydrate complexes in solution by NMR. Although different types of clasification could be envisaged we have decided to distinguish between those methods which permit us to deduce (a) structural information on the protein residues involved in the interaction; (b) information on the bound carbohydrate; and (c) information on the entire protein– carbohydrate complex.135 Chemical Society Reviews 1998 volume 27 Table 2 Summary of the studies described in the text showing the method used by their authors System WGA/Sialic acid Hevein/GlcNAc-containing oligosaccharides UDA/GlcNAc-containing oligosaccharides Ac-AMP2/GlcNAc-containing oligosaccharides CBD C. fimi Cen C/cello oligosaccharides CBD T. Reesei/cello oligosaccharides Tripeptides/heparin disaccharide Macrophage mannose receptor/mannose Hevein pseudohevein UDA WGA/chitooligosaccharides Ricin-B/galactose derivatives IgA X24 antibody/fluorinated disaccharide Galectin/galactose disaccharide AAA/fucose disaccharide Antibody/Salmonella trisaccharide Ricin/C-lactose Strep-9 antibody/GlcNAc-containing trisaccharide E-selectin/SLeX P-,L-selectin/SLeX Dolichos biflorus lectin/Blood group A trisaccharide Dolichos biflorus lectin/Forsmann pentasaccharide Lentil lectin/sucrose Acidic fibroblast growth factor/sucrose octasulfate A.niger glucoamylase/N S-heteroanalogues of maltose Polyclonal IgG/GD1a ganglioside CBD T. Reesei/cello oligosaccharides Immunoglobulin G glycoforms (glycoprotein) HCGa glycans (glycoprotein) Hevein/GlcNAc-containing oligosaccharides CBD C. fimi/cello oligosaccharides Verotoxin VT-1/globotriaosyl ceramide RNAse B glycoforms (glycoprotein) Adhesion domain of human CD2 (glycoprotein) Human granulocyte colony stimulating factor (glycoprotein) Fucosylated peptide (glycoprotein) HCG a subunit (glycoprotein) 2.1 Methods which allow us to obtain information on the protein residues involved in the interaction 2.1.1 Titration experiments NMR is the prevalent method used to study molecular conformation and dynamics in solution in contrast to X-ray crystallography methods.NMR spectroscopy provides detailed information on the chemical surroundings of a given nucleus through chemical shifts and therefore titration NMR experiments may provide an adequate means to analyse sugar-induced perturbations of proteins and vice versa. Usually the specific binding of carbohydrates to a lectin is monitored by recording the 1H NMR spectra of a series of samples with variable sugar concentration (six to eleven different concentrations).The concentration of the protein during the experiments is held constant. A first sample is used to obtain the 1H NMR chemical shifts of the free-sugar lectin sample (dfree). A second sample is prepared by dissolving a large amount of the corresponding sugar in a similar protein solution. The titration curve is built by adding small aliquots of the sugar-concentrated protein solution over the free-sugar protein sample. This protocol is a way to verify the existence of complexes between a given lectin and the corresponding carbohydrates and in addition the alterations in the chemical shifts of the amino acid proton resonances may be used to determine the equilibrium association constant K Chemical Society Reviews 1998 volume 27 136 Ref.Method 8 9 10 Titration Titration Titration 11 12 13 14 16 15 Titration Titration Titration Titration Titration 17 20 19 21 22 25 27 CIDNP TR-NOESY TR-NOESY TR-NOESY TR-NOESY TR-NOESY TR-NOESY 18 18 23 TR-NOESY TR-NOESY TR-NOESY 6 TR-NOESY 24 18 TR-NOESY TR-NOESY 26 TR-NOESY 1 TR-NOESY 18 Relaxation 33 Relaxation 14 Relaxation 1 Relaxation 29 30 complete 3D 9 complete 3D 12 13 complete 3D 35 complete 3D complete 3D 18 complete 3D 31 complete 3D 32 complete 3D 29 30 a considering the equilibrium eqn.(1)–(3) where dbound corresponds to (1) Protein + Sugar/?Protein sugar Protein × Sugar] (2) = Ka [ [ Protein] � [Sugar] (3) ) d - (d + d = d free bound free obs [Protein × Sugar] [Protein] + [Protein × Sugar] Scheme 5 Structure of N-acetyl glucosamine (GlcNAc) containing sugars. From top to bottom GlcNAc chitobiose chitotriose. × the NMR chemical shifts of the sugar-bound form of the lectin. The values of Ka and dbound may be obtained by non-linear leastsquares fitting of the observed NMR chemical shifts dobs of different selected protons of the lectin as a function of the total sugar concentration. dfree may be introduced as an adjustable parameter to control the goodness of fit by comparing its value with that obtained experimentally.Moreover provided that the protein resonances have been assigned this methodology may be used to locate at least qualitatively the sugar binding site around certain protein residues. Thus either qualitative or quantitative estimations of binding affinity and specificity can be derived by using 1D or 2D spectroscopy. Kinetic and thermodynamic parameters may also be inferred at least qualitatively by running these titration experiments at different temperatures and following a van’t Hoff type of analysis using a representation of R log Ka vs. 1/T. 2.1.1.1 GlcNAc-binding proteins The pioneering work in this field was performed by Kronis and Carver,8 who analysed the interaction and thermodynamics of the binding of sialyl oligosaccharides by the lectin wheat germ agglutinin (WGA).As a recent example of the application of this protocol Asensio et al.9 have recently reported the determination of the binding site of hevein by using NMR spectroscopy and different N-acetyl glucosamine-derived (GlcNAc) ligands. The GlcNAc chitobiose and chitotriose (Sche 5) specific binding constants were also determined by 1D NMR spectroscopy. These constants increase by one order of magnitude when passing from the mono- to the di- and to the tri-saccharide. In addition the thermodynamic parameters for chitotriose–hevein and chitobiose–hevein interactions were obtained from a van’t Hoff analysis indicating that the association process is enthalpy driven while entropy opposes binding.This behaviour is usually observed for protein–carbohydrate interactions.1–3 The deduced negative signs indicated that hydrogen bonding and/or van der Waals forces are the major interactions stabilizing the complex. The differences in binding constants were explained in terms of the three-dimensional structure of the complexes, also obtained from NOESY NMR spectroscopy (see below in Section 2.3). A similar study has also been performed to deduce the chitotriose-induced perturbations in Urtica dioica agglutinin (UDA) which contains two homologous hevein domains. The data confirmed the presence of two binding sites of nonidentical affinities since sugar induced perturbations occur in one domain of the lectin at sugar concentrations below equimolar.Residues in the second domain are shifted at higher trisaccharide concentrations.10 The interaction between chitotriose and a related antifungal and antimicrobial peptide Ac- AMP-2 has also been studied by 1H NMR showing that as observed for hevein and UDA three aromatic residues are involved in binding.11 2.1.1.2 Cellulose-binding proteins The binding specificity of the interaction of different glucans with the 152 amino acid N-terminal cellulose binding domain (CBD) of Cellulomonas fimi CenC has also been studied by 1H NMR showing that at least four b-(1?4) linked glucopyranosides (Scheme 6) are Scheme 6 Structure of b-(1?4)-linked glucose oligomers n = 2 cellotetraose n = 3 cellohexaose required to detect binding.12 The cellulose binding domain spans five glucosyl units.Using an NMR model of the protein (see below in Section 2.3) it was deduced that the interaction takes place primarily through hydrogen bonding and van der Waals stacking. Titration experiments of a related binding domain with cellohexaose allowed it to be established that Trp54 and Trp72 participate in cellulose binding. Using an NMR derived structure of this polypeptide it was deduced that both residues are adjacent in space and exposed to solvent forming a ligand binding cleft which is a feature common to the cellulose binding domains of the same family.13 The identification of the functionally relevant amino acids of the cellulose binding domain from Trichoderma reesei cellobiohydrolase I has been performed by recording the NMR spectra of synthetically modified peptides.Although in general the structural effects of substitutions were minor in some cases decreased binding could clearly be ascribed to conformational perturbations. At least one glutamine and two tyrosine residues were found to be essential for tight binding (see below in Section 2.3).14 2.1.1.3 Other examples One recent example of the application of titration NMR experiments has focused on the mechanism of calcium and sugar binding to a C-type carbohydrate recognition domain of the macrophage mannose receptor. The authors varied the nature of several key amino acid residues of the protein using site directed mutagenesis. Titration NMR experiments were performed for every mutant and from the corresponding magnitudes of the affinity constants it was possible to deduce that a stacking interaction between mannose and Tyr729 contributes about 25% of the total free energy of binding.15 13C NMR has also been used for titration experiments.For instance the interaction of the heparin disaccharide with tripeptides has been studied by titration and 13C NMR relaxation measurements. Relaxation rates (see below under 2.2.2) for the disaccharide are significantly higher in the presence of the peptide. The analysis of the data in terms of molecular diffusion constants indicated that the peptide is oriented proximal to the uronic acid ring.16 2.1.2 Chemically induced dynamic nuclear polarization (CIDNP) In many cases sugar recognition by proteins involves the side chains of tryptophan tyrosine and histidine moieties (Scheme 2) so the photo CIDNP method may profitably be applied to monitoring the effect of ligand binding on the receptor.These aromatic moieties are able to produce CIDNP signals after laser irradiation in the presence of a suitable radical pair-generating dye. Elicitation of such a response in lectins implies accessibility of the respective aromatic groups to the light-absorbing dye. Therefore this protocol may be suitable for monitoring surface properties of a protein receptor and the effect of sugar binding provided that CIDNP-active amino acid residues are involved in the recognition site. Experimentally the intensity and the shape of the CIDNP signals are therefore determined in the absence and in the presence of different carbohydrate ligands.This method has been elegantly applied recently by Siebert et al. to study the complexation of GlcNAc-containing oligosaccharides to a series of plant lectins of increasing structural complexity.17 In particular the binding of chitosugars (Scheme 5) to hevein pseudohevein Urtica dioica agglutinin (UDA) wheat germ aglutinin and its B domain were investigated. When the sugar is bound CIDNP signals of the aromatic moieties of Tyr Trp or His are altered with respect to those of the free protein they may be broadened appear with reduced intensity or even disappear completely. Thus their involvement in sugar binding may be deduced. The results obtained were in agreement with those previously reported.9,10 In addition a conformational change of an indole ring of a Trp residue was also detected for UDA.2.2 Methods which allow us to obtain structural information on the bound carbohydrate 2.2.1 Transferred nuclear Overhauser enhancement (TR-NOE) studies It is obvious that knowledge of the recognised conformation of a biologically active carbohydrate presents considerable implications for rational drug design. From the three dimensional point of view TR-NOE may allow the assessment of the conformation of protein-bound oligosaccharides.18 NOESY experiments provide information about which protons are close in space and therefore they may be used to deduce conformational information. The TR-NOESY is a regular NOESY experiment but it is applied to a protein–ligand system in dynamic exchange in which the ligand is present in excess.For ligands which are not bound tightly and exchange with the free form at a reasonably fast rate as usually observed for carbohydrates TR-NOE provides an adequate means to determine their bound conformation (Scheme 7). In complexes involving large molecules cross relaxation rates of the bound compound (sB) are opposite in sign to those of the free one (sF) and produce negative NOEs. Therefore the existence of binding may be easily deduced by visual inspection since NOEs for small molecules are positive (Scheme 8). The conditions for the applicability of this approach are well established considering the well known equilibrium and the molar fractions of free and protein bound sugar eqn.(4)–(6) (4) Protein + Sugar [ K = (2) a [ (excess)/?Protein sugar Protein × Sugar] Protein] � [Sugar] p K (5) (6) bsB > pfsF 1 >> sB 21 is the off-rate constant. where pb and pf are the fractions of bound and free sugar ligand and sB and sF the cross relaxation rates for the bound and the free ligand respectively. K Under these conditions it can be considered that eqn. (7) holds. 137 Chemical Society Reviews 1998 volume 27 1 2 (7) sobs = pbsB + pfsF 3 Scheme 7 Schematic representation of the facts which take place during TR-NOESY experiments. (1) Initial state free sugar. (2) Formation of the complex bound sugar.NOEs between protons close in space are developed. (3) Dissociation of the complex. The free sur maintains the information acquired in the bound state for a given period of time which depends on its relaxation times. After this time the system reverts to the initial state. TR-NOESY experiments are usually performed at different mixing times and ligand–protein ratios and produce strong negative NOEs on ligand binding. However one of the major drawbacks of this experiment is the possible existence of spin diffusion effects which are typical for large molecules. In this case apart from direct enhancements between protons close in space other spins may mediate the exchange of magnetization thus producing negative cross peaks between protons far apart NOESY ac d d d c c b b a a RELAXATION 138 b Scheme 8 Left.Schematic representation of a NOESY spectrum for a free sugar. Cross peaks and diagonal peaks have different signs. Right. Schematic representation of a TR-NOESY spectrum recorded for an exchanging sugar–protein system. Cross peaks and diagonal peaks have the same signs as expected for a large molecule thus indicating binding to the protein. The relative sizes of the peaks and the appearance of new ones may be used to detect conformational variations. Chemical Society Reviews 1998 volume 27 in the macromolecule. Thus protein-mediated indirect TRNOE effects may lead to interpretation errors in the analysis of the ligand bound conformation. As a prime example one of the first reported applications of TR-NOE experiments was to the derivation of oligosaccharide bound conformations which concluded that a fluorinated Gal-b-(1?6)-Gal-b-OMe disaccharide underwent major conformational changes around the glycosidic linkages when bound to a specific antibody.The conclusion was based on the detection of an NOE cross peak between two protons located on two different pyranoid moieties. However the reevaluation of the problem by the same authors using TR-NOEs in the rotating frame (TR-ROESY) experiments19 demonstrated that this cross peak was dominated by an indirect effect mediated by a protein proton. In TRROESY spin-diffusion (three spin) effects appear as positive cross peaks and therefore the application of this experiment permits one to distinguish direct from indirect enhancements and thus complements those measured under regular conditions providing conformational information which is less contaminated by artifacts (Scheme 9).Different examples which have focused on the study of the structural events that mediate the molecular recognition processes between proteins and carbohydrates have recently been presented through examples of lectin- antibody- and enzyme-type receptors. Specific comparisons of the differential binding of natural and modified analogues have also been reported. 2.2.1.1 Recognition of the global minimum conformation Although there is not any general rule in many cases protein binding sites are well preorganized to recognise a conformation of the oligosaccharide which is located close to the its global minimum energy region.Several TR-NOE studies on protein– carbohydrate interactions have studied the complexes between Ricin-B chain and different oligosaccharides.20 Ricin is a TR-NOESY a bc d a c b a c b d d c c b b a a TR-ROESY Scheme 9 Schematic representation of TR-ROESY (left) and TR-NOESY (right) spectra for an exchanging sugar–protein system. In TR-ROESY spin diffusion (three spin) cross peaks (i.e. a/c) and diagonal peaks have the same signs. On the other hand direct cross peaks (a/b) show different sign to diagonal peaks (a). In TR-NOESY all direct and spin diffusion-mediated cross peaks have the same sign as diagonal peaks. dimeric (A and B chains) galactose-binding lectin seeds which has been shown to be 10–100 fold more toxic to some transformed cell lines than to normal cells and has therefore been considered as a potential antitumor agent.The first NMR study of Ricin-B/disaccharide complexes used monodimensional (1D)-TR-NOE experiments to study the binding of Ricin-B by methyl b-lactoside. In this example it Scheme 10 Structure of different galactose-containing disaccharides and analogues TR-NOESY was demonstrated by using a selectively deuterated substrate that there were minor changes in the conformation of free methyl b-lactoside (Scheme 10) upon binding to Ricin-B. In a second example the Ricin-B-bound conformation of melibiose [Gal-b(1?6)-Glc] was deduced and compared to its conformation in free solution demonstrating that only one of the two solution conformations of melibiose was recognized by the Chemical Society Reviews 1998 volume 27 139 lectin.Therefore the protein causes a shift in the solution equilibrium towards the bound conformation during the recognition process. Docking studies indicated that the protein chain excluded binding of certain ligand conformations on the basis of unfavorable interactions between the protein surface and remote parts of the carbohydrate. However since Ricin-B preferentially binds b-galactosides rather than a-galactosides and since the orientation of the glucose residue in a-substituted galactosyl glucosides is very different from that existing in their b-analogues the conclusions reached for the melibioside could not be extrapolated in a general way.Thus in an attempt to generalise this structural problem the conformational changes that occur when methyl a-lactoside was bound to the Ricin-B chain in aqueous solution were then studied.20 The observed data indicated that the protein causes a slight conformational variation in the glycosidic torsion angles of methyl a-lactoside although the recognized conformer was still within the lowest energy region. Molecular modeling using molecular dynamics minimization and docking of the disaccharide within the binding site of Ricin B strongly suggested that apart from the expected contacts between the galactose moiety and different amino acid residues there were also van der Waals contacts between the protein and the remote glucose moiety as previously deduced from binding studies using modified lactoside derivatives.Thus both van der Waals contacts and hydrogen bonding contribute to the stability of the complex.20 As with Ricin–lactose there are other cases in which there are no major variations in the conformational behavior of the oligosaccharide upon protein binding. For instance the TRNOESY study of the binding of Galb-(1?2)Galb-(1?R) to the galectin of chicken liver showed that the conformation of the disaccharide in the bound state is very close to its global energy minimum state in solution.21 2.2.1.2 Simultaneous recognition of different conformations There are examples in which the protein does not select a single conformer.The Aurelia aurantia agglutinin (AAA) lectin recognises simultaneously different conformations22 of Fuca( 1?6)GlcNAcb-(1?OMe). This disaccharide which is fairly flexible when free in solution appears to remain to a certain extent flexible around the glycosidic linkage within the lectin binding site. An analogous case has been reported for the complex between methyl b-allolactoside [Galb(1?6)Glcb- OMe] and Ricin B.20 In this case and contrary to the observations for lactose different conformations around the f y and w glycosidic bonds of methyl b-allolactoside were recognized by the lectin. In fact for this complex only the TRNOESY cross-peaks corresponding to the protons of the galactose residue were negative as expected for a molecule in the slow motion regime.In contrast the corresponding cross peaks for the glucose residue were ca. zero as expected for a molecule whose motion is practically independent of the protein. 2.2.1.3 Protein-induced conformational selections Lectins and antibodies may select just one of the conformers present in the conformational equilibrium for the free state. The quest for the active conformation of the Lewis X oligosaccharide has stimulated different research groups. The sialyl Lewis X (SLeX {aNeuNAc-(1?3)b-Gal(1?4)[aFuc-(1?3)]Glc} tetrasaccharide exists in solution as an equilibrium of several conformations which are mainly characterized by the orientations of the N-acetylneuraminic acid residue. Perhaps the pre-eminent study on this topic has been reported by Peters and coworkers18 using spin-locked filtered NOESY and Metropolis Monte Carlo calculations.The most relevant conclusion is that E-selectin complexes exclusively to a conformation of sialyl Lewis from the conformational equilibrium in aqueous solution in which the sialic acid shows an orientation defined by (f/y 76/6) already reported to be present in free solution. On the other hand the orientation of the fucose residue (f,y 38/26) differs from that preferred in aqueous solution. This work clarified previous discussions on the conformational changes of sLeX upon Chemical Society Reviews 1998 volume 27 140 binding to E-selectin. Recently the bound conformation of sLeX bound to E-selectin was also compared to those recognised by P- and L-selectin.23 In all cases it was demonstrated that the conformation of the branched trisaccharide remained close to the conformation of the free ligand.However E- and P-selectins recognised a different conformation around the sialic acid glycosidic linkage than L-selectin. The blood group A trisaccharide {aGalNAc-(1?3)[aFuc- (1?2)]b-Gal} exists in solution as an equilibrium between two families of low energy conformers. Comparison between experimental and simulated TR-NOESY volumes lead to the conclusion that only one conformation of the trisaccharide was bound6 by the GalNAc-specific lectin isolated from Dolichos biflorus. Spin diffusion NOEs were detected by means of TRROESY experiments. The proposed bound conformation was in agreement with one of the two deduced from previous modeling studies.As with other lectins complementary forces emanate from hydrogen bonding and van der Waals forces including hydrophobic interactions. A second report on the application of TR-NOE experiments to the molecular recognition of oligosaccharides by the seed lectin of Dolichos biflorus has been completed.24 TR-NOESY and TR-ROESY experiments collected for a mixture of this lectin and the Forssman pentasaccharide GalNAca-(1?3)Gal- NAcb-(1?3)Gala-(1?4)Galb-(1?4)Glc revealed close contacts between the non-reducing disaccharide moiety of the carbohydrate and the lectin binding site. In addition and using an elegant protocol of recording experiments at different lectin:sugar ratios the authors deduced two distinct classes of NOE cross peaks which reflected the size of the carbohydrate epitope and thus also of the binding pocket of the lectin.In order to detect contacts between the protein and the carbohydrate chain T2-filtered TR-NOESY spectra were performed which permitted the detection of NOEs between the terminal disaccharide fragment and protein protons most likely belonging to Leu residues in agreement with the previously reported molecular modeling study of the complex. 2.2.1.4 Protein-induced conformational variations Several cases of protein-induced major conformational changes have also been reported. Bundle and coworkers25 have presented TRNOESY evidences which show that a branched trisaccharide {a-Galp(1?2)[a-Abep(1?3)]-Manp-1?OMe} related to the antigenic determinant of a Salmonella polysaccharide undergoes an antibody-induced conformational shift about one glycosidic linkage (Gal-Man) when bound in solution.Previous data have demonstrated that only this trisaccharide portion of the complete polysaccharide was bound by the antibody. Although the TR-NOESY distance constraints were compatible with two different bound conformations one of them was shown to be consistent with the X-ray structure of the same molecular complex but none with the free solution conformation of the oligosaccharide. The Strep 9 antibody-bound conformation of a branched trisaccharide namely GlcNAc-b-(1?3)-a-Rha- (1?2)-a-Rha-OMe has been18 investigated by TR-NOESY and TR-ROESY experiments and Metropolis Monte Carlo calculations.It was deduced that the monoclonal antibody Strep 9 selects only one defined conformation of the carbohydrate hapten. This bound conformation which is a local energy minimum on the potential energy maps of the free ligand undergoes a change in the orientation of one glycosidic linkage when compared to the global minimum conformation in the free solution state. It was also necessary to include repulsive constraints derived from the absence of NOEs to deduce the three dimensional structure of the trisaccharide in the binding site of the antibody. The conformational features of one of the most important of the food industry-relevant sugars sucrose in the combining site of lentil lectin in solution have been recently characterized through TR-NOESY experiments and molecular modeling.18 The experimental NMR data which indicated that the bound sucrose has a unique conformation for the glycosidic linkage were in agreement with the results obtained for the complex using X-ray crystallography.It is important to mention that major differences with respect to the hydrogen bonding network of free sucrose were found since none of the two inter-residue hydrogen bonds detected in crystalline sucrose were conserved in the complex with the lectin. Stacking interactions between a Phe residue and the hydrophobic face of the glucose residue as well as between the same Phe and H-4 and H-6 of the fructose moiety were deduced both experimentally (X-ray) and by modeling.A variety of protein–sugar hydrogen bonds were also detected. In addition the NMR study provided insight into the residual conformational flexibility of sucrose in the lectin binding site. On the other hand it has also been shown that free sucrose octasulfate appears to assume a conformation significantly different from any of the X-ray conformations determined for sucrose when bound to the acidic fibroblast growth factor. In this case strong electrostatic interactions between guest and host may be the dominant factor in the deformation of sucrose octasulfate.26 2.2.1.5 Use of structurally modified carbohydrate analogues Obviously not only natural ligands but also structurallymodified oligosaccharides may be used as lectin ligands or as inhibitors of carbohydrate-processing enzymes and thus these analogues may be employed to deduce enzymatic mechanisms.Moreover conformational differences between free and protein-bound natural carbohydrates and synthetic analogues may also be assessed by TR-NOE experiments. Ricin B has also been used as a model to study the bound conformation of potential glycosidase inhibitors such as Cglycosides. 27 Although many reports have usually assumed that the conformation of free C-glycosides was the same as that of the corresponding O-analogues it has recently been reported that at least for O- and C-lactoses this is not the case.27 Thus 2D TR NOESY experiments were recorded to study the complexation of C-lactose by Ricin B. The conformational study of C-lactose in the free state showed that the exo-anomeric conformation around the C-glycosidic bond was adopted.However the conformation around the aglyconic bond was rather different to that of the natural compound. For O-lactose ca. 90% of the population was located around the so called minimum syn f/y 54/18 and ca. 10% of population around minimum anti f/y 36/180.20 However C-lactose was shown to exhibit much higher flexibility than its O-analogue and three conformational regions (syn anti and gauche–gauche) were significantly populated in solution (Scheme 4). The comparison between the NOESY and ROESY spectra of C-lactose recorded in the absence and in the presence of the lectin indicated that conformers syn and gauche–gauche were not bound.Therefore the experimental results indicated that Ricin B selects different conformers of C-lactose (anti) and its O-analogue (syn).20 In order to estimate the relative binding affinities of the C- and O-glycosides competitive TR-NOEs with different O-lactose/C-lactose ratios were also performed. It was demonstrated that both ligands compete for the same binding sites of the lectin and that the affinity constant of C-lactose is smaller than that of its O-analogue. Although merely speculative and since the flexibility of C-lactose in the free state is much higher than that of O-lactose the cause of the recognition of different conformations could be of entropic origin. Other modified carbohydrates have been employed to study the structure of enzyme–inhibitor complexes.Mario Pinto and coworkers18 have investigated A. Niger glucoamylase. This enzyme catalyzes the hydrolysis of maltose-type molecules with inversion of configuration. In an elegant manner and using novel heteroanalogues of maltose containing sulfur in the nonreducing ring and nitrogen in the interglycosidic linkage they have recently demonstrated that methyl 5A-thio-4-N-a-maltoside is a potent enzyme inhibitor and that it is bound by the enzyme in a conformation close to its global minimum. The characteristic NOEs observed for a second conformer which is also present in free solution as a minor form were not detected in the presence of glucoamylase G1. It is noteworthy that the crystal structure of a complex of a closely related glucoamylase with dihydroglucoacarbose indicated that the bound conformation of this ligand resembles the global minimum and that an existing local minimum conformer cannot be readily accommodated by the enzyme because of adverse van der Waals interactions.2.2.1.6 Other developments TR-NOESY experiments have also been recently applied to the identification and characterization of biologically active molecules from a mixture.28 As already stated above the sign of transferred NOEs is opposite to that of NOEs of small molecules that do not bind to a protein and thus an unequivocal and fast identification of molecules with binding properties is possible. 2.2.2 NMR relaxation measurements NMR relaxation properties depend on the spectral density functions [J(w)] which in turn are sensitive to molecular motion.Spectrometer frequency molecular size and internuclear distances are also important parameters. In fact spectral densities are related to the motional correlation times (tc) which measure the rate of molecular tumbling. The functional form of the basic longitudinal (T1) and transversal (T2) relaxation times are given below in eqn. (8)–(11). (8) T121 = (W/20) [J(wH2wc) + 3J(wc) + 6J(wH + wc)] (9) T221 = (W/20) [4J(0) + J(wH2wc) + 3J(wc) + 6J(wH) + 6J(wH + wc)] 2tc J(nw) = (10) 2 1 + (n c (11) wt ) C H 2pr3 CH æ g g hö ø ÷ W = N è ç Exchange and dynamic processes affect heavily these relaxation parameters. Since T2 values are directly related to linewidths (T2* = 1/pu1/2) the simple measurement or estimation of linewidths (u1/2) may serve as a basis to deduce the occurrence of a dynamic process (such as binding or recognition) in the vicinity of a given nucleus (Fig.1). Therefore for instance provided that the NMR signals of the protein have been assigned 15N and/or 13C NMR relaxation measurements at one or several magnetic fields may be used to probe the change of mobility of specific amino acid residues of carbohydraterecognizing proteins upon sugar binding. However to the best of our knowledge this aproach has not yet been used. Nevertheless it is obvious that the relaxation properties of the oligosaccharide will also be affected upon protein binding due to their dependence on molecular motion.Therefore provided that the 1H or 13C NMR signals of the sugar have been assigned NMR relaxation measurements may be used to probe the change of mobility of specific carbohydrate moieties upon binding. As an example of this application the interaction between cellohexaose and cellulose binding domains from Trichoderma reesei cellulases has been studied by T2 relaxation analysis of the oligosaccharide resonances. In addition and using an NMR derived structure of the polypeptide a model for the molecular complex has been proposed in which three aligned aromatic residues (tyrosines) with a precise spatial arrangement stack onto every other glucose ring of the cellulose polymer.14 The glycoproteins approach has recently been used by Dwek and coworkers1 to probe the relative mobility of different glycoforms of immunoglobulin G observing that mobility is dependent on the primary sequence of the glycan.Homans and coworkers have analysed line widths and 1H and 13C chemical shift changes of the glycan moieties on isotopically 13C 15N Chemical Society Reviews 1998 volume 27 141 Fig. 1 Comparison between the 1H NMR spectrum recorded for a free disaccharide (top) and that recorded for the same sugar in the presence of a specific lectin (bottom). In the case of the spectrum of the bottom part (sugar/protein 20/1 ratio) the change in linewidths indicates that an interaction is taking place since the relaxation properties of the disaccharide are heavily affected in the bound state.enriched recombinant human chorionic gonadotropin a-subunit.29 They have shown that the biologically relevant glycan at Asn 52 appears to extend into solution both in the isolated a subunit and in the complex with the b subunit. Similar conclusions have been deduced by the group at Utrecht in this case using a natural abundance sample.30 Other studies of the motion of the glycan moieties in different glycoproteins have been performed using either relaxation measurements or line width analysis. These include the adhesion domain of human CD2,18 the human granulocyte-colony-stimulating factor, 31 a fucosylated peptide,32 several glycoforms of RNAseB.1 In all cases the overall mobility of the glycan chain is reduced as compared to that of the corresponding small glycopeptides or oligosaccharides.The results obtained also suggest that the carbohydrate moiety reduces the local mobility around the glycosylation site. In addition the carbohydrate provides more kinetic and thermodynamic thermal stability to the protein. The influence of 9-O-acetylation of GD1a ganglioside on the recognition by a natural human antibody has been analysed by molecular dynamics simulations and NMR.33 Although acetylation did not influence the overall conformation of the ganglioside the NMR spectrum of the acetylated GD1a in the presence of the polyclonal IgG showed the disappearance of the 9-O-acetyl signal indicating a variation in the value of T2 thus its involvement in an exchange process and consequently that the interaction with the human protein takes place on this site.2.3 Complete determination of the 3D structure of protein–carbohydrate complexes In a few favorable cases dealing with protein receptors small enough to be amenable to direct analysis to NMR methods 1H NMR techniques have been used to deduce the three dimensional structure of protein–carbohydrate complexes. Very recently and following the impressive development in NMR methodology and in molecular biology methods (allowing the obtention of 15N- and 13C-labelled molecules) detailed structural information on the 3D structure of carbohydrate–protein complexes and glycoproteins in solution has become available Chemical Society Reviews 1998 volume 27 142 by using modern NMR pulse sequences especially designed for NMR active heteroatoms.29 Hevein9 is a protein of 43 amino acids whose structure has independently been solved by X-ray at 0.28 nm resolution and by NMR methods.Interestingly although the structure of hevein in water–dioxane and water solutions differs significantly from that observed in the crystal it closely resembles the solid state structures of the domains of wheat germ agglutinin (WGA). Asensio et al.9 have recently reported on the determination of the structure of the complex of hevein with chitobiose (Scheme 5 see above in section 2.1.1.1) by using NMR spectroscopy. Using NOESY spectroscopy and restrained molecular dynamics they also presented a refined NMR structure of free hevein in water. The structure of the complex of hevein with methyl b-chitobiose has also been derived recently (Asensio et al.in press). Protein–carbohydrate NOEs measured for the hevein–chitobiose and hevein–methyl b-chitobiose complexes (Asensio et al. submitted) allowed the deduction of the conformation of these complexes. Obviously the presence of NOESY cross peaks between certain protons of the sugar and the protein permit one to infer that these atoms are close in space and therefore to derive the three dimensional structure of the complex. No important changes in the protein NOEs were observed indicating that carbohydrate-induced conformational changes in the protein are small. The N-acetyl methyl signal of the non-reducing GlcNAc moiety of b-chitobiose displayed NOE contacts with Tyr30 and Trp21 residues and appeared strongly shielded.From the inspection of the model a hydrogen bond between Ser19 and the non-reducing N-acetyl carbonyl group was suggested as well as one between Tyr30 and HO-3 of the same sugar residue. The previously mentioned N-acetyl methyl group of the non-reducing GlcNAc displayed non polar contacts to the aromatic Tyr30 and Trp21 residues. Moreover the higher affinities deduced for the b-linked oligosaccharides with respect to GlcNAc and GlcNAca-(1?6)-Man could be explained by favorable stacking of the second b-linked GlcNAc moiety and Trp21. The final 3D structures derived by NMR were compared to those of WGA Ac-AMP II (which is also a GlcNAc-binding protein) recently solved by NMR Martins et al.34 and to the crystal structure of hevein.The corresponding average rmsd are 0.060 nm (B domain of WGA residues 16–32) 0.100 nm (Ac-AMP2 residues 12–32) and 0.269 nm (crystal of hevein residues 16–41). The structure of the 152 amino acid N-terminal cellulose binding domain of Cellulomonas fimi CenC has also been derived by multidimensional heteronuclear NMR in the presence of saturating concentrations of cellotetraose (Scheme 6).12 The polypeptide is composed of ten beta strands folded into two antiparallel beta sheets with the topology of a jellyroll beta sandwich. These strands form the face of the protein previously determined by titration experiments to be responsible for cellulose binding (see above in Section 2.1).There is a binding cleft which contains a central strip of hydrophobic residues that is flanked on both sides by polar hydrogen bonding groups. The existence of this cleft provides a structural explanation for the selectivity of this binding domain. Recently the solution structure of the carbohydrate-binding B-subunit homopentamer of verotoxin VT-1 from E. coli complexed to globotriaosylceramide [aGal-(1?3)bGal- (1?4)bGlc?R] has been deduced by Homans and coworkers, 35 using a 13C/15N doubly labelled protein sample. Unlike the crystal structure in solution there is no evidence of anomalous association between two of the subunits which may be an artifact of crystallization. TR-NOEs obtained for the complex compare satisfactorily with those predicted from a previous molecular modeling study of the complex.Although they are not protein–carbohydrate complexes it has to be noted that recently several glycoprotein structures in solution along with their structure-related properties have also been derived by NMR. These have been mentioned in Section 2.2.2. and are the adhesion domain of human CD2,18 the human granulocyte-colony-stimulating factor,31 a fucosylated peptide, 32 several glycoforms of RNAseB,1 and human chorionic gonadotropin.29,30 3 Perspectives With no doubt present and future developments including expression systems for glycoproteins,29 will allow us to produce these biomolecules in the required amounts for detailed analysis of NMR data. Current NMR methodologies,29,35 which permit one to deduce dynamic parameters through relaxation measurements will also find their application in the derivation of differential flexibility of the protein binding site before and after complex formation.A similar methodology may be applied to the ligand molecule provided it is also 13C-labelled. Moreover the use of new methods to detect long lived protein-bound water molecules by NMR in combination with other biophysical techniques will surely allow us to dissect the relative contribution of van der Waals hydrogen bond water-mediated and entropy contributions to the stabilization of the carbohydrate –protein supramolecule. 4 Acknowledgments We thank Drs J. L. Asensio J. F. Espinosa M. Mart�ýn-Pastor J. Ca�nada M. Bernab�e and M.Mart�ýn-Lomas for helpful discussions and their contributions. We also thank SIdI-UAM for the facilities provided and DGICYT (PB96-0833) and the Mizutani Glycoscience Foundation for financial support. 5 References 1 R. A. Dwek Chem. Rev. 1996 96 683. 2 W. I. Weis and K. Drickamer Ann. Rev. Biochem. 1996 65 441. 3 E. Toone Curr. Opin. Struct. Biol. 1994 4 719 and references therein. 4 J. M. Rini Curr. Opin. Struct. Biol. 1995 5 617 and references therein. 5 A. Imberty K. D. Hardman J. P. Carver and S. Perez Glycobiology 1991 1 456. 6 F. Casset T. Peters M. Etzler E. Korchagina N. Nifant’ev S. Perez and A. Imberty Eur. J. Biochem. 1996 239 710. 7 R. J. Woods Curr. Opin. Struct. Biol. 1995 5 591 and references therein. 8 K. A. Kronis and J.P. Carver Biochemistry 1985 24 826. 9 J. L. Asensio F. J. Ca�nada M. Bruix A. Rodriguez-Romero and J. Jimenez-Barbero Eur. J. Biochem. 1995 230 621 and references therein. 10 K. Hom M. Gochin W. J. Peumans and N. Shine FEBS Lett. 1995 361 157. 11 P. Verheyden J. Pletinckx D. Maes H. A. M. Pepermans L. Wyns R. Willem and J. Martins FEBS Lett. 1995 370 245. 12 P. E. Johnson M. D. Joshi P. Tomme D. G. Kilburn and L. P. McIntosh Biochemistry 1996 35 14381 and references therein. 13 G. Y. Xu E. Ong N. R. Gilkes D. G. Kilburn D. R. Muhandiram M. Harris-Brandts J. P. Carver L. E. Kay and T. S. Harvey Biochemistry 1995 34 6993. 14 M. L. Mattinen M. Kontelli J. Kerovuo M. Linder A. Annila G. Lindeberg T. Reinikainen and T. Drakenberg Protein Sci.1997 6 J. F. G. Vliegenthart C. S. Wright and H. J. Gabius Proteins 1997 28 294 and references therein. 15 N. P. Mullin P. G. Hitchen and M. E. Taylor J. Biol. Chem. 1997 272 5668. 16 D. Mikhailov K. H. Mayo A. Pervin and R. J. Lindhardt Biochem J 1996 315 447. 17 H. C. Siebert C. W. Von der Lieth R. Kaptein J. J. Beintema K. Dijkstra N. van Nuland U. M. Soedjanaamadja A. Rice 268. 18 T. Peters and B. M. Pinto Curr. Opin. Struct. Biol. 1996 6 710 and references therein. 19 S. R. Arepalli C. P. J. Glaudemans D. G. Davis P. Kovac and A. Bax J. Magn. Reson. B 1995 106 195 and references therein. 20 J. L. Asensio F. J. Ca�nada and J. Jimenez-Barbero Eur. J. Biochem. 1995 233 618 and references therein. 21 H. C. Siebert M. Guilleron H. Kaltner C. W. Von der Lieth T. Kozar N. Bovin E. Y. Korchagina J. F. G. Vliegenthart and H. J. Gabius Biochem. Biophys. Res. Comm. 1996 219 205. 22 T. Weimar and T. Peters Angew. Chem. Int. Ed. Engl. 1994 33 88. 23 L. Poppe G. S. Brown J. S. Philo P. V. Nikrad and B. H. Shah J. Am. Chem. Soc. 1997 119 1727 and references therein. 24 F. Casset A. Imberty S. Perez M. Etzler H. Paulsen and T. Peters Eur. J. Biochem. 1997 244 242. 25 D. R. Bundle H. Baumann H. Brisson S. Gagne A. Zdanov and M. Cygler Biochemistry 1994 33 5183. 26 J. Shen and L. Lerner Carbohydr Res. 1995 273 115. 27 J. F. Espinosa J. Ca�nada J. L. Asensio M. Mart�ýn-Pastor H. J. Dietrich R. R. Schmidt M. Mart�ýn-Lomas and J. Jim�enez-Barbero J. Am. Chem. Soc. 1996 118 10682 and references therein. 28 B. Meyer T. Weimar and T. Peters Eur. J. Biochem. 1997 246 705. 29 C. T. Weller J. Lustbader K. Seshadri J. M. Brown C. A. Chadwick C. E. Kolthoff S. Ramnarain S. Pollak R. Canfield and S. W. Homans Biochemistry 1996 35 8815. 30 T. de Beer C. W. E. M. van Zuylen B. R. Leeflang K. Hard R. Boelens R. Kaptein J. P. Kamerling and J. F. G. Vliegenthart Eur. J. Biochem. 1996 241 229. 31 V. Gervais A. Zerial and H. Oschkinat Eur. J. Biochem. 1997 247 286. 32 G. Mer H. Hietter and J. F. Lefevre Nature3 45. 33 H. C. Siebert C. W. Von der Lieth X. Dong G. Reuter R. Schauer H. J. Gabius and J. F. G. Vliegenthart Glycobiology 1996 6 561. 34 J. Martins D. Maes R. Loris H. A. M. Pepermans L. Wyns R. Willem and P. Verheyden J. Mol. Biol. 1996 258 322. 35 J. M. Richardson P. D. Evans S. W. Homans and A. Donohue-Rolfe Nature Struct. Biol. 1997 4 190. Received 14th August 1997 Accepted 7th November 1997 143 Chemical Society Reviews 1998 volume 27

ISSN:0306-0012    

DOI:10.1039/a827133z

出版商:RSC    

年代:1998

数据来源: RSC

摘要:

The aza-Payne rearrangement a synthetically valuable equilibration Toshiro Ibuka Graduate School of Pharmaceutical Sciences Kyoto University Sakyo-ku Kyoto 606-01 Japan An aza-Payne rearrangement of N-activated 2-aziridinemethanols synthesized in an optically active form with potassium tert-butoxide (ButOK) potassium hydride (KH) or sodium hydride (NaH) at near 0 °C in common aprotic solvents such as tetrahydrofuran (THF) toluene or a mixed solvent of THF–HMPA followed by quenching at low temperature gives the corresponding epoxy sulfonamides in high yields. The anionic reaction intermediates generated by treatment of 2-aziridinemethanols with a base react readily in a one-pot manner with a variety of nucleophiles such as organocopper reagents thiols and trimethylsilyl cyanide to yield the optically active corresponding functionalized 1,2-amino alcohols in good yields.Upon exposure of 2,3-epoxy amines to an equimolar mixture of ButOK and BuLi (super base) in a mixed solvent of THF and n-hexane at 278 °C the equilibrium lies exclusively toward the hydroxyaziridine forming direction. 1 Introduction Historically the first documented report of a possible Payne rearrangement in carbohydrate chemistry was probably that of British chemists Lake and Peat who reported around 60 years ago the isomerization of one epoxy alcohol into another under basic conditions.1 The next major advance in terms of this type of isomerization however came in 1957 when Angyal and Gilham demonstrated unambiguously that the hydroxy group on the C-1 carbon atom adjacent to the epoxide ring in 1 could carry out an intramolecular nucleophilic displacement yielding the more stable epoxy alcohol 2 and this process was referred to as ‘epoxide migration’ (Scheme 1).2 This epoxide migration has become a well-known reaction in the area of carbohydrate chemistry.3 Toshiro Ibuka was born in Japan in 1936.He graduated from Gifu Pharmaceutical University in 1960 and obtained MSc and DrSc degrees in 1962 and 1965 from the Faculty of Pharmaceutical Sciences Kyoto University. He joined the faculty at Kyoto University in 1965 and is currently Professor of Organic Chemistry. He has spent a year and three months as a visiting professor at the University of North Carolina and three months as an exchange scientist at the Centre de Neurochimie- CNRS Strasbourg.His current research interest is focused on the development of stereoselective synthetic methods by the use of organometallic reagents and their application to the synthesis of biologically important molecules. He has published more than 160 research papers and reviews and several books. X X HN HN H Nu– Nu R1 R1 R2 O R2 OH H 10 11 OH O 3 3 2 C C C C C C1 1 2 O HO 1 2 H OH HO O OH O base Nu OH R1 R1 R1 2 3 1 3 1 2 Nu– H R2 R2 R2 5 3 N2 type mechanism. 4 Scheme 1 The most influential paper on this type of reaction however was Payne’s 1962 publication describing the equilibration (epoxide migration) of racemic simple 2,3-epoxy alcohols under basic conditions.4 In the case of the intramolecular nucleophilic displacement of simple 2,3-epoxy alcohols such as 3 the term ‘Payne rearrangement’5,6 has been widely adopted in honor of this pioneering work rather than the wording ‘epoxide migration’.One important aspect of the rearrangement is that the reaction is usually stereospecific proceeding with inversion of configuration at the C-2 carbon of the epoxide-ring via an S Payne also reported the considerable synthetic potential of the rearrangement although this potential was apparently overlooked by the chemists of the time. In addition because the relative proportions of epoxy alcohols 3 and their isomeric terminal epoxides 4 at equilibrium are highly structure dependent,4 the preparative value of the Payne rearrangement under aqueous alkaline conditions has long been rather limited.Since the discovery of the Sharpless asymmetric epoxidation of allylic alcohols to prepare chiral 2,3-epoxy alcohols 3,7 the regioselective nucleophilic ring-opening at the C-1 position of more reactive terminal epoxides 4 resulting from the Payne rearrangement of 2,3-epoxy alcohols 3 in basic media has become an important tool for the synthesis of polyfunctional synthetic building blocks 5 (Scheme 1). Such a one-pot transformation has been termed a ‘Payne rearrangement–ring opening reaction’ by Sharpless.5 On the other hand aziridines are often vital structural units of biologically active molecules and are frequently employed as synthetic building blocks of natural compounds and elsewhere.8 Recently by using Lewis acids such as TMSOTf N,Ndibenzylated or N,N-diallylated 2,3-epoxy amines compounds of type 6 have been reported by Rayner and co-workers to rearrange to the aziridinium salts 7 which open at the more reactive C-1 position with various nucleophiles to yield the rearrangement–opening products 8.9–11 For example the 2,3-epoxy-N,N-diallylamine derivative 6 (R1 = Prn R2 = H) was treated with TMSOTf to yield the aziridinium salt 7 in quantitative yield.Removal of the trimethylsilyl group of 7 with potassium carbonate in methanol and subsequent intramolecular rearrangement regenerated the original epoxy amine starting material 6 in a high yield.It should also be stressed that the rearrangement of the epoxy amines 6 to the aziridinium salts 7 is irreversible (Scheme 2). For a more complete overview of Chemical Society Reviews 1998 volume 27 145 O R2 3 R1 R1 NR3 2 H 6 HN X R2 N OH R1 O R2 H H or tosyl etc. 2 R1 Me 3 H 9 H N Ts R1 OH 14 NH2 Me Me Ts—N H 12 Route A X MeCu(CN)Li Route B 1 MeCu(CN)Li (81% yield) Scheme 2 R1 R2 = alkyl or aryl; R3 = allyl or benzyl X = CO2R mesyl the above described methods developed by Rayner to obtain enantiomerically pure compounds the reader may refer to a recent and excellent review.11 3,14 BunLi–Me3Al,15 Ti(OPri)4,16 and SiO2.17 Until recently due to its reversible nature the aza-Payne rearrangement of 2-aziridinemethanols 9 and 2,3-epoxy amines of type 12 has been much less extensively examined (Scheme 2).12–19 Although the relative proportions of the 2-aziridinemethanols 9 and the epoxy amines 10 would be substrate as well as reaction condition dependent conflicting experimental evidence for the direction of the aza-Payne rearrangement has been reported.Some reactions of epoxy sulfonamides 10 (X = Ts) have been shown to produce 2-aziridinemethanols of type 9 under basic conditions.12 In contrast with some N-tosyl 2-aziridinemethanols 9 the equilibrium favours the epoxy N-tosylamides 10.13 Furthermore epoxy amines of type 12 provide access to hydroxyaziridines 13 in the presence of Lewis acids such as BF During a study of organocopper reactions (Scheme 3),18 we found that the reaction of 2-aziridinemethanol 14 with methylcyanocuprate [MeCu(CN)Li·LiBr] afforded two products.To our surprise the products were neither the expected primary alcohol 15 nor the isomeric primary alcohol 16 (Route A) but the secondary alcohols 17 and 18. The formation of the secondary alcohol 17 suggested a reaction path proceeding via the aza-Payne rearrangement–ring opening (Route B).19 13 OH + Me + (58 42) Scheme 3 There has been no systematic investigation of the abovedescribed phenomena and the issue of delineating the factors which are responsible for controlling the direction of the aza- Payne rearrangement remains a topic of much attention.We therefore examined the equilibrium of the aza-Payne rearrangement by chemical and theoretical studies in simple systems.19 It is not the purpose of this review to present detailed information of all available preparative methods for the requisite substrates nor to give an encyclopaedic discussion of the aza-Payne rearrangement. Rather I wish to present an Chemical Society Reviews 1998 volume 27 17 146 OTMS TMSO H 1 2 Nu– Nu R1 R2 R2 N H R3 R3 +N R3 R3 7 8 X X HN H Nu– R1 O R2 OH H OH R2 H 10 H Me R1 NHTs Nu OH Me R2 H H H 11 Me H 16 OH H 18 NH Me Ts—N H Me Ts—N 15 OH H overview of recent advances made by our team and other groups that may be important in the understanding of this chemical reaction and to fill some of the voids described above by studying the potential of the aza-Payne rearrangement reaction.2 Synthesis of 2-aziridinemethanols Various N-substituted or unsubstituted enantiomerically pure (or enriched) 2-aziridinemethanols may be prepared by two broadly applicable methods viz. by starting from enantiomerically pure 2,3-epoxy alcohols20,21 which in turn could be readily prepared from allylic alcohols by the use of the Sharpless asymmetric epoxidation,7 or by starting from synthetic or natural amino acids.20,22 Because the reactivity of NHaziridines towards nucleophiles is low activation by the introduction of an electron-withdrawing group on the ring nitrogen atom is required.The term ‘activated aziridines’ has been introduced by Ham for aziridines that easily undergo nucleophilic SN2-type ring-opening in the absence of a formal positive charge on the nitrogen atom.23 The arenesulfonyl methanesulfonyl acyl and alkoxycarbonyl moieties serve as good activating groups while nucleophilic ring-opening reactions of aziridines involving a positively charged nitrogen have been extensively studied by Rayner.11 2.1 Synthesis from 2,3-epoxy alcohols In terms of availability and cost of synthetic precursors chiral 2,3-epoxy alcohols which are readily prepared from suitable allylic alcohols by the use of Sharpless asymmetric epoxidation, 7 are by far the most common starting materials for the synthesis of various types of 2-aziridinemethanols.The example shown in Scheme 4 illustrates a synthetic route to the 2-aziridinemethanol 25 starting from the chiral 2,3-epoxy alcohol 19 (Scheme 4).20 OTBS 3 OTBS OH O BnO i + (71%) BnO OH N BnO 3 OR (49 51) 21 19 R = H 20 R = TBS 22 ii iii BnO OTBS BnO OH N iv H H H H N N (74%) R Ts 25 23 R = H 24 R = Ts (46% from 22) 2Si; Ts = 3–NH4Cl refluxing H2O–2-me- 3 in refluxing THF; iii Et3N–TsCl in THF; iv Bu4NF Scheme 4 Abbreviations Bn = benzyl; TBS = But(Me) 4-methylphenylsulfonyl. Reagents i NaN thoxyethanol; ii PPh in THF. The silyloxy epoxide 20 prepared from 2,3-epoxy alcohol 197 (86% ee) was reacted with sodium azide in the presence of ammonium chloride24 to yield a ca.1 1 mixture of two products 21 and 22.20 Although the separation of these two isomers could be conveniently accomplished by flash silica gel chromatography the reaction of a mixture of 21 and 22 with triphenylphosphine followed by tosylation yielded the N-tosyl aziridine 24 via 23 as the sole isolable product which upon exposure to tetrabutylammonium fluoride yielded the target N-activated 2-aziridinemethanol 25. Although the enantiomeric excess (ee) of the starting epoxy alcohol 19 was only 86% essentially enantiopure 2-aziridinemethanol 25 ( > 98% ee) was obtained by recrystallization of the crude product20 and many other 2-aziridinemethanols could be prepared in a similar way. 2.2 Synthesis from amino acids Synthesis of N-activated 2-aziridinemethanols can also be achieved starting from chiral natural or synthetic amino acids such as l-threonine 26 and d-allo-threonine 31 as outlined in Scheme 5.N-Trityl 2,3-cis-aziridine 27,25 readily prepared in Me Me OH CO2Me OH 2 3 3 2 H H H H Me iv v i - iii 1 1 N N NH CO2H 2 R Tr 26 (84%) (61%) (32%) Me OH Me OH CO2Me 28 R = Tr 29 R = Ts 30 R = Ms 3 2 iv H H Me H H i - iii 1 1 N N NH (58%) CO2H 2 Tr Tr 33 31 vi (65%) Me 32 3 2 OH Me CO2Me vii iv H Me H 27 3 2 1 NHTs N (90%) NHTs 36 34 3 Ts 5 (58%) OH iv Me OR OH OTBS OH OH viii CO2Me vii H H Me Me N (83%) ix NHTs NHTs Ts 38 37 (87%) (95%) 39 R = TBS 14 R = H Scheme 5 Abbreviations Tr = triphenylmethyl; Ts = 4-methylphenylsulfonyl; Ms = methanesulfonyl; TBS = tert-buthyldimethylsilyl.2–MeOH then TsCl–Et3N; vii PPh3–diethyl azodicarboxylate; viii Reagents i SOCl2–MeOH; ii TrCl–Et3N in DMF iii MsCl–pyridine then reflux; iv DIBAL-H; v TFA then TsCl–Et3N or TFA then MsCl–Pri 2NEt; vi SOCl TBSCL–imidazole–4-DMAP; ix Bu4NF. high yields from l-threonine 26 was then reduced with diisobutylaluminum hydride (DIBAL) to yield the 2-aziridinemethanol 28 ( > 99% ee) in high yield. Removal of the trityl group followed by tosylation (or mesylation) gave N-tosyl (or N-mesyl) 2-aziridinemethanol 29 ( > 98% ee) or 30 ( > 98% ee).20 Synthesis of the isomeric aziridine 2,3-trans 14 proved considerably more troublesome and various standard routes were attempted unsuccessfully.Thus d-allo-threonine 31 while readily converting into N-tritylaziridine ester 32 failed to give the desired 2-aziridinemethanol 33 on reduction with DIBAL a surprising observation in view of the successful reduction of the isomeric ester 27 under the same conditions. Next reduction of the readily available N-tosyl aziridine 35 was examined. The aziridine methyl ester 35 derived from d-allo-threonine 31 via the N-tosyl methyl ester 34 in the usual way was treated with DIBAL to yield a mixture of products from which the undesired ring-opened product 36 was isolated as the major product. Other reducing agents were tried but without success. Finally the silyloxyaziridine 39 was easily synthesized from the N-tosyl ester 34.Selective silylation of diol 37 which in turn was derived from 34 by reduction with DIBAL and the subsequent aziridine-ring formation yielded the 2-aziridinemethanol silyl ether 39 which upon desilylation with tetrabutylammonium fluoride afforded the desired N-tosyl 2-aziridinemethanol 14 ( > 98% de).20 Synthetic methods for the preparation of some N-activated 2-aziridinemethanols from appropriate amino acids such as lserine d-serine d-threonine and l-methionine have also appeared recently.20,22 2.3 Asymmetric synthesis Recently elegant asymmetric syntheses of N-activated 2-aziridinemethanols have been described26,27 and these two methods can be applied to the synthesis of various types of synthetically important 2-aziridinemethanols bearing a phenyl group.The major limitation of these methods is that only non-enolizable N-trimethylsilyl imines like 43 or sulfinimines 47 can be prepared from aromatic aldehydes efficiently. Me Me Me B Br Me Me Me PhCH 42 Ph 40 3 (80%) H NH2 Br S N—SiMe 43 O p-Tolyl Ph N 47 Scheme 6 Reagents i LiAlH4; ii TsCl–Et3N Treatment of the chiral bromoborane 40 and tert-butyl bromothioacetate 41 in the presence of triethylamine yields the boron enolate 42 which on reaction with the trimethylsilyl imine 43 yields a-bromo-b-amino thioester 44 in a very high diastereo- and enantio-selective manner. Reduction of 44 with lithium aluminum hydride then affords the 2-aziridinemethanol 45 which can easily be converted into the N-activated 2-aziridinemethanol 46.26 Another highly diastereoselective one-pot synthesis of N-activated 2,3-cis-2-aziridinemethanols has been accomplished by the use of a Darzens-type reaction.27 Thus exposure of the lithium enolate 48 of methyl bromoacetate to optically pure sulfinimine 47 yields 2,3-cis-aziridine ester 49 in high yield with the added advantage that the N-sulfinyl can be removed under mild conditions.Interestingly both removal of the N-sulfinimine group and reduction of the ester group in 49 could be carried out by treatment with LiAlH4 to yield directly chiral 2-aziridinemethanol 45.27 Evans and co-workers have developed an efficient asymmetric aziridination reaction (Scheme 7).28 Reaction of methyl cinnamate 50 with [N-(p-tolylsulfonyl)imino]iodobenzene 51 in the presence of 5 mol% of a catalyst derived from copper(i) triflate and chiral 4,4A-disubstituted bis(oxazoline) 52 afffords the aziridine ester 53 with high enantioselectivity.Both CuOTf and Cu(OTf)2 can be employed in the formation of competent catalysts and similar enantioselectivities were observed in each case. The transition metal-catalyzed transfer of diazocarbonylderived carbenes to imines has been studied by using the N-benzylidene aniline 54 by Jacobsen and his co-workers.29 Jacobsen found that the copper(i) complex [CuPF6(MeCN)4] in the presence of the ligand 52 was most effective in catalyzing Chemical Society Reviews 1998 volume 27 49 Br SBut BrCH2COSBut O *L B 42 COSBut Ph OH 41 i (86%) H H 1 ii (89%) L* 3 2 N R i Ph Br OMe OLi 45 R = H 46 R = Ts CO2Me 3 2 H H 44 48 N (65%) p-Tolyl-S O (2 S,3 S) 147 PhI NTs Ph CO2Me CuOTf Ligand 52 50 O N Ph Ph N N Ph 52 Ph N H N 51 Ph O OTs H 58 O 54 2CHCO2Et 55 CuPF6(MeCN)4 Ligand 52 H Prn H Bun OTs H 60 56 ee 44% Scheme 7 the aziridination reaction to yield aziridines 56 and 57 which like esters 53 could be transformed into the corresponding 2-aziridinemethanols by reduction.Although more than a dozen other attractive and excellent methods for the synthesis of chiral 2-aziridinemethanols and their analogues have been developed, 30 detailed descriptions of all synthetic methods for the requisite 2-aziridinemethanols are beyond the scope of this article.3 Synthesis of 2,3-epoxy amine substrates Although there are many methods available for the synthesis of epoxides possessing internal secondary or tertiary amino groups Rayner synthesized several terminal N,N-diallyl- or -dibenzyl-2,3-epoxy amines from the corresponding tosylates of 2,3-epoxy alcohols by sodium iodide-catalyzed displacement using N,N-diallylamine or N,N-dibenzylamine.9–11 A typical example is illustrated by the transformation of the tosylate 58 into the N,N-disubstituted terminal amine 59 (Scheme 8). KI HNBn2 (80%) i NaN3 (77%) ii PPh3 (80%) Scheme 8 Epoxides like 61 which possess a primary amino group at the terminal position can easily be synthesized from the corresponding 2,3-epoxy alcohol tosylates like 60 according to a two-step sequence of reactions.15 This involves (1) substitution of the tosyloxy group by treatment with sodium azide and (2) reduction of the azide group with triphenylphosphine.31 However one should appreciate that low molecular weight epoxy amines such as 61 pose serious problems with respect to isolation as the usual extractive workup leads to considerable loss of the water-soluble product.Consequently after the reduction of azides with triphenylphosphine is complete the reaction mixture is usually concentrated and then the product may be distilled using a bulb-to-bulb distillation apparatus under reduced pressure.In this way 2,3-epoxy amines are easily obtained in pure form. 4 Aza-Payne rearrangement reaction of N-activated 2-aziridinemethanols As has been mentioned earlier in this article the Payne rearrangement of 2,3-epoxy alcohols is usually carried out in Chemical Society Reviews 1998 volume 27 148 Ph CO2Me H H N Ts 53 O (yield 63%) (ee 94%) Ph Ph CO2Me CO2Me H H + H N Ph (4 1) (37% yield) 57 ee 35% O H Prn NBn2 H 59 H O Bun NH2 H 61 aqueous sodium hydroxide and results in an equilibrium mixture of the starting 2,3-epoxy alcohol and the isomeric 3-hydroxy-1,2-epoxide.4–6 However there have been conflicting publications about the favoured direction of the corresponding aza-Payne reaction.Whereas reaction of N-tosyl epoxy sulfonamides 63 65 and 67 has been reported to yield 2-aziridinemethanols 62 64 and 66 respectively by brief heating in aqueous basic conditions,12 with 2-aziridinemethanol 68 the equilibrium lies exclusively toward the opposite direction to yield N-tosyl epoxy sulfonamide 6913 (Scheme 9). NH—Ts Ts N aq. NaOH H O OH 63 62 (90%) NH—Ts Ts N Me aq. NaOH Me O OH 65 64 (80%) NH—Ts Ts N H aq. NaOH O Me OH Me Me Me 67 66 (93%) OH H H O H N NHTs TBDPSO Ph H N Ts 70 HO TBDPSO H Ph O Ts OH N Ts LiOBu t HMPA THF aq. NaOH ButOH (61 39) aq. NaOH ButOH (2 98) Scheme 10 H 72 69 (95%) H O NH—Ts 71 H NHTs 73 68 Scheme 9 Abbrevations TBDPS = tert-butyldiphenylsilyl In the case of N-Boc or N-trityl aziridines poor yields of rearrangement products were also obtained from base-catalyzed isomerization reactions and consequently we initiated our study on the aza-Payne rearrangement to determine the scope of the reaction with respect to reaction conditions.Not unexpectedly exposure of N-tosyl-2-aziridinemethanol 70 to an aqueous sodium hydroxide solution in the presence of tert-butyl alcohol as co-solvent at 0 °C for 18 h produced a 61 39 equilibrium mixture of 70 and a rearranged product 71 (Scheme 10). Much higher selectivity (72 73 = 2 98) is found with H H activated aziridine 72 in which there is branching at the hydroxy-bearing carbon.From the results shown in Scheme 10 it is apparent that the relative proportions of the 2-aziridinemethanols (70 72) and epoxy sulfonamides (71 73) at equilibrium are highly structure dependent. Consequently the preparative value of the aza-Payne rearrangement under aqueous alkaline conditions is limited. In addition the aqueous conditions employed preclude the use of many synthetically important organometallic nucleophiles such as organocopper reagents for subsequent reactions in a one-pot manner. The present review also illustrates how this problem can be overcome. Because knowledge of the relative thermodynamic stability of N-activated 2-aziridinemethanols and 2,3-epoxy amines is the key to an understanding of the isomerization phenomenon we have also undertaken ab initio molecular orbital calculations in simple compounds.The energy minimum of N-mesyl- 2-aziridinemethanol 74 was predicted to be only 1.7 kcal mol21 higher than the energy minimum of epoxy sulfonamide 75 (Scheme 11).19 OH O HN H N 74 H HM MeO2S H 75 3O+ H SO2Me HM 3O+ O– H O N H N SO2Me 77 MeO2S Scheme 11 76 Next under basic conditions in an aprotic solvent would the oxa-anion 76 of N-mesyl-2-aziridinemethanol 74 or the azaanion 77 of epoxy sulfonamide 75 be expected to be the more stable species? Both theoretical and experimental aspects of this study were carried out with N-mesyl-2-aziridinemethanol 74 and epoxy sulfonamide 75 and the geometries of optimized reactant oxa-anion 76-A the transition state TS involved in the rearrangement step and the product aza-anion 77-A as well as relative energies are shown in Fig.1. Fig. 1 RHF/3-21+G* optimized geometries and their relative energies for the aza-Payne rearrangement TS the transition structure; 76-A the reactant oxa anionic energy minimum; 77-A the product aza anionic energy minimum. We found that the reactant oxa-anionic energy minimum 76-A was predicted to be ca. 16.22 kcal mol21 higher in energy than the product aza-anionic energy minimum 77-A. Consequently exclusive formation of the epoxy sulfonamide 75 could OH aq. NaOH H H O 3 2 Me (30 70) H H 1 N NH—Ts Ts 78 Me Entry Base Yield of 78 (%) 29 KH (4 equiv.) KH (4 equiv.) KH (4 equiv.) KH (4 equiv.) KH (4 equiv.) KH (4 equiv.) 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 a 1.KH (4 equiv.) in THF 2. aq. NH4Cl (0 100) Scheme 12 be expected by treatment of 74 with a base such as ButOK potassium hydride (KH) sodium hydride (NaH) or lithium diisopropylamide (LDA) followed by quenching at low temperature (Scheme 11). In fact the N-mesyl-2-aziridinemethanol 74 did give only the rearranged product 75 upon exposure to NaH (1.3 equiv.) in a mixed solvent of THF–HMPA (12 1) followed by quenching at 278 °C and we were unable to detect unreacted starting 74 by 1H NMR analysis.19 As can be seen from Scheme 12 exposure of 2-aziridinemethanol 29 to aqueous NaOH provides a 30 70 equilibrium mixture of the aziridine 29 and the rearranged product 78.In order to identify suitable reaction conditions for this rearrangement the influence of base solvent and reaction temperature were studied in more detail using the readily available N-tosyl-2-aziridinemethanol 29 as the test starting material (Table 1). Based on this study we can offer some general comments on factors that influence the successful outcome of the aza-Payne reaction.19 Table 1 Aza-Payne rearrangement of (2S,3S)-3-methyl-N-tosyl-2-aziridinemethanol 29 to yield epoxy sulfonamide 78a Conditions Solvent THF–HMPAb RT (3 h) 0 °C (18 h) RT (1 h) 278 °C (5 h) 220 °C (4 h) 0 °C (5 h) RT (2 h) THF–HMPAb THF THF THF THF THF THF–HMPAd RT (2 h) 0 °C (2 h) 220 °C (2 h) 0 °C (30 min) DBU (5 equiv.) BunLi (1.2 equiv.) LDA (4 equiv.) KH (4 equiv.) NaH (4 equiv.) NaH (4 equiv.) NaH (4 equiv.) NaH (4 equiv.) ButOK (1.2 equiv.) THF–HMPAb THF THF toluene CH2Cl2 DME 0 °C (1 h) 0 °C (1 h) 0 °C (2 h) RT (30 min) 1,4-dioxane 0 17 0c < 1 < 1 77 82 92 99 96 96 99 99 99 99 Isolated and unoptimized yields of epoxy sulfonamide 78.b THF– HMPA = 10 1. c A complex mixture of products was obtained. d THF– HMPA = 12 1. (1) In the case of N-Boc or N-trityl aziridines poor yields were obtained from base-catalyzed isomerization reactions however this problem can be overcome by the use of N-alkylor N-aryl-sulfonylated aziridines.(2) Although DBU (1,8-diazabicyclo[5.4.0]undec-7-ene) BuLi and LDA were inappropriate for clean and efficient rearrangements (Entries 1–3 Table 1) bases such as NaH ButOK and KH gave satisfactory results (Entries 8–15 Table 1). Using KH as a base the rate of rearrangement of activated aziridinemethanols was much faster than that observed using NaH under otherwise identical conditions. For example treatment of 29 with NaH at 0 °C for 5 h yields the epoxy sulfonamide 78 in 77% yield while exposure of the same substrate to KH at 0 °C for 30 min affords 78 in 96% isolated yield (entries 6 and 11 Table 1). (3) The reaction temperature is an important factor and hence the temperature should be carefully controlled.For example while 29 reacted with KH at 220 °C after 2 h to afford only the rearranged product 78 in 96% isolated yield (entry 10) treatment of the same substrate 29 at 278 °C for 5 h led to complete recovery of the unchanged starting material (entry 4). Chemical Society Reviews 1998 volume 27 149 Although the rearrangement rates are dependent upon the base used the reaction normally proceeds completely in the direction of the epoxy sulfonamide 78 at either 0 °C or RT (ca. 25 °C). (4) THF or a mixture of THF and HMPA is the solvent of choice. Although rearrangement of 29 with NaH in THF alone is rather slow (entries 5 6 and 7) acceleration of the rearrangement is accomplished by the addition of HMPA (entry 8).Other common solvents such as toluene dichloromethane dimethoxyethane and 1,4-dioxane can be used equally well for the reaction of 29 with KH (entries 12–15). The bases NaH and KH were found to be the best among those tested for the rearrangement. As can be seen from Scheme 12 although exposure of the aziridine 29 to an aqueous sodium hydroxide solution gave a 30 70 mixture of 29 and 78 treatment of 29 in THF with KH gave exclusively 78. Thus we have used these two bases for the reaction of some representative 2-aziridinemethanols. Satisfactory results obtained for 5 different activated aziridines (14 25 30 79 and 80) bearing a primary hydroxymethyl group at the C-2 position of the aziridine-ring are summarized in Scheme 13.By exposure to either NaH or KH followed by quenching all aziridinemethanols yielded the corresponding rearranged epoxy sulfonamides 81–85 in high isolated yields.19 OH Me H O H KH (4 equiv.) 0 oC 30 min Me H H N (92%) NH—Ts 81 Ts 14 BnO OH H O H H H NaH (4 equiv.) RT 2 h BnO NH—Ts ( 97%) N Ts 25 82 OH Me H O H Me NaH (1.5 equiv.) 0 oC 2 h H H ( 80%) NH—Ms N Ms 83 30 OH H O HN H Ts NaH (4 equiv.) RT 1 h ( 65%) N Ts 84 79 H Ph OH O H KH (4 equiv.) Ph 0 oC 1.5 h H H NH—Ts 91% 85 N Ts 80 Scheme 13 All reactions were carried out in THF–HMPA (10 ~ 12 1). Abbreviations Ts = 4-methylphenylsulfonyl; Ms = methanesfulfonyl; Bn = benzyl.In a similar manner as shown in Scheme 14 activated aziridines (86 87 and 72) having a secondary or tertiary hydroxy group react equally well to yield the respective epoxy sulfonamides (88 89 and 73) in high yields. In all the rearrangements listed in Schemes 13 and 14 no evidence for unrearranged starting material was detected by HPLC analysis of the crude reaction product(s). Thus the reaction appears to be quite general for activated aziridines possessing wide structural variety giving isolated yields which are good to excellent. In addition the reversible nature of the reaction in protic solvents as shown in Scheme 10 can be altered to an essentially irreversible process in aprotic solvents.19 It should be clearly noted that the aziridine 72 (Scheme 14) possessing a tertiary hydroxy group rearranged slowly in THF in the presence of KH to give solely the epoxy sulfonamide 73.On the other hand similar treatment with KH (4 equiv.) but in CH2Cl2 at room temperature for 18 h unexpectedly provided a Chemical Society Reviews 1998 volume 27 150 H O H KH (4 equiv.) H N Ts Me NHTs Me CH2Cl2 0 oC 4 h (77%) H OH 88 86 H O H H KH (4 equiv.) N Ts Bun NHTs Bun THF 0 oC 2 h (88%) H OH 89 87 H HO Ts N O H KH (4 equiv.) NHTs THF RT 18 h (99%) 73 72 Ts Ts N N H H O O N2 H H H H O OH O O i 90 Scheme 14 mixture of 73 (85% yield) and an undesired compound 90 (ca. 5% yield) rather than pure 73. Consequently despite the slow rate of rearrangement in THF this is the solvent of choice for the reaction of compounds of type 72 bearing a tertiary hydroxy group.19 Formation of the minor product 90 could be explained in the following manner.In the first step reaction between CH2Cl2 and KH will produce a carbene species (:CHCl). Trapping of the carbene by an alkoxide generated by exposure of the alcohol 72 to KH will produce a chloromethyl ether which will react with another alkoxide under typical S conditions to yield the product 90. Formation of the major product 73 is an example of an intramolecular SN2 reaction and presumably would be much faster than the side reaction to form 90. Not unexpectedly the attempted rearrangement reaction of N-tosyl-(S)-azetidinemethanol 91 or N-tosyl-(S)-prolinol 92 with NaH or KH in THF led to recovery of the starting material.The reaction in CH2Cl2 under otherwise identical conditions gave solely the corresponding adducts 95 and 96 in high yields (Scheme 15). This demonstrates that the rearrangement reactions are specific to three-membered aza-cycles possessing a large strain energy.19 ii N N N Ts Ts NHTs Ts 95 91 93 H H H O H O O OH i iii N N NHTs N Ts Ts Ts 96 94 92 Scheme 15 Reagents and conditions i KH in THF 0 °C 0%; ii KH in CH2Cl2 0 °C 3 h 84%; iii KH in CH2Cl2 0 °C 3 h 95% 5 Aza-Payne rearrangement reaction of 2,3-epoxy amines The direction of the equilibrium of the aza-Payne rearrangement of epoxy amines has been investigated by many research groups.9–17,19 It is well established by Rayner9–11 that quantitative generation of quaternary aziridinium salt 98 could be accomplished by exposure of 2,3-epoxy N,N-diallylamine 97 to trimethysilyl triflate (Scheme 16); a process that is essentially OTMS H O TMSOTf TfO– Prn Prn N N H + H H 98 K2CO3 MeOH Scheme 16 97 irreversible.However desilylation of the aziridinium salt 98 by treatment with potassium carbonate in methanol regenerates the original 2,3-epoxy diallylamine 97 and it would appear that at least in this present case desilylation and subsequent aziridinium ion ring-opening to yield original 2,3-epoxy amine 97 is a more favourable process. In other words 2,3-epoxy amine 97 is thermodynamically more stable than the corresponding aziridinium ion.Would a 2,3-epoxy amine or its isomeric 2-aziridinemethanol be expected to be the more thermodynamically stable compound? In order to reduce the size of the problem such that ab initio calculations could be employed model systems of 2,3-epoxy amine 99 and 2-aziridinemethanol 100 were chosen for study. We found that the energy minimum 99-A of 2,3-epoxy amine 99 was predicted to be ca. 4.77 kcal mol21 lower in energy than the energy minimum 100-A of 2-aziridinemethanol 100 at the RHF/3-21G level (Fig. 2).19 Thus calculations predict that the 2,3-epoxy amine 99 is more stable rather than the 2-aziridinemethanol 100 and based on these calculations we were apprehensive as to the possibility of successfully isomerizing 2,3-epoxy amines such as 99 into 2-aziridinemethanols of type 100.Fig. 2 RHF/3-21G optimized structures 99-A of 99 and 100-A of 100 Several research groups14–17 studied the aza-Payne rearrangement on various types of 2,3-epoxy amines including amino sugars and typical examples are listed in Scheme 17. It is apparent that all reactions were carried out in the presence of a Lewis acid such as BF3·Et2O BunLi–Me3Al Ti(OPri)4 or silica gel. Voelter reported that the reaction of the amino sugar 101 in Et2O with BF3·Et2O in the presence of trimethylsilyl azide produced the sugar 105 involving an aziridine ring group.14 Vaultier and co-workers found that the exposure of 2,3-epoxy amine 102 to a mixture of BunLi and Me3Al in THF furnished the 2-aziridinemethanol 106.Both BunLi and Me3Al are essential for a clean transformation.15 Sato and co-workers discovered that Ti(OPri)4 is an effective catalyst for the aza- Payne rearrangement of 2,3-epoxy amine 103 into the aziridine alcohol 107.16 Interestingly Voelter and co-workers reported that epoxy amines like 104 afforded the corresponding aziridine alcohols of type 108 merely by exposure to silica gel in a mixed solvent of n-hexane and EtOAc.17 Thus under Lewis acidic conditions the direction of the rearrangement of epoxy amines 101–104 to yield the corresponding aziridine alcohols 105–108 has to be considered as a general rule. We have also studied the rearrangement of 2,3-epoxy amines but under basic conditions. A number of common bases such as metal hydrides (NaH KH etc.) and metal alkoxides (ButOK OBn O O NHR 101 R = C Ph NH2 6H4CO2But 1.Bu nLi H H O 102 H2N O Ph OH 103 OBn O O NHR 104 NH Me H Bu NH2 H O 109 H O O 110 2 H 3 H 1 111 Me O Me R = CH(Me)CO2Bn Scheme 17 Abbreviation Bn = benzyl EtONa etc.) were reacted unsuccessfully with epoxy amines 109 and 110. The rearrangement of epoxy amines with MeLi and BunLi proceeds very slowly but this reaction does not go to completion. For example although the epoxy amine 110 reacted with MeLi and BunLi to yield the hydroxyaziridine 115 the reaction never proceeded beyond 83% and 71% completion respectively (Scheme 18). 2 2 2 –78 oC 1 h (85%) H Me O NH NH 2 Me Me 2 NH 1 112 3 H 113 Scheme 18 All reactions were carried out in n-hexane–THF (1 2 ~ 1 5) with a mixed reagent of BunLi/ButOK (1 1) (1.5 ~ 2.0 equiv.) After further study we did find that treatment of the 2,3-epoxy amines 109 and 110 with a mixture of 1.5 equiv.of BunLi and 1.5 equiv. of ButOK (‘super base’)32 at 278 °C to yield only the corresponding rearranged hydroxyaziridines 114 and 115 respectively in high yields with no detectable traces of Chemical Society Reviews 1998 volume 27 118 OBn HO Me3SiN3 O BF3·Et2O (53%) N R 105 N H Ph 2. Me3Al 3. NaF OH (73%) 106 OH Ti(OPri)4 Ph THF 0 oC N OH (63%) H 107 OBn HO O SiO2 (75%) N R 108 H H H N Me OH 114 H H H N Bu –78 oC 1.5 h OH (85%) 115 H H N HO –78 oC 1.5 h (96%) 116 OH Me Me –78 oC 1.5 h Me (88%) 117 Me N H H N Me 0 oC 3 h H (40%) OH 151 the corresponding reactants 109 and 110.It should be noted that these low molecular weight water-soluble hydroxyaziridines such as 114 also pose a problem with respect to product isolation (see above). Consequently following the reaction of the epoxy amine 109 with the super base the reaction was quenched at 278 °C with saturated aqueous ammonium chloride. The mixture was allowed to warm to 0 °C then filtered through a short Celite pad. Concentration followed by recrystallization or distillation using a bulb-to-bulb distillation apparatus under reduced pressure yielded the pure material 114.Disubstitution at the C-3 position of 2,3-epoxy amines such as 111 and 112 does not exert any influence on the rearrangement giving aziridines 116 and 117 respectively. As anticipated increasing the steric bulk at the C-2 carbon in 113 does decrease the relative reaction rate for the aza-Payne rearrangement and yields the expected aziridine alcohol 118 in rather low yield (compare the reaction of 111 and 112 with 113 Scheme 18). The study demonstrates that the aza-Payne reaction of 2,3-epoxy amines with ButOK/BunLi proceeds at 278 °C in high isolated yields except for the less reactive substrate 113. 6 One-pot aza-Payne rearrangement-epoxide ring opening reaction of 2-aziridinemethanols.Simple stereoselective synthetic route to synthetically important 1,2-amino alcohols Acyclic 1,2-amino alcohols play an important role as chiral auxiliaries and chiral building blocks in the preparation of biologically active compounds.33 For example enantiomerically pure N-tosyl-1,2-amino alcohol 119 was treated with triallylborane to yield B-allyloxazaborolidine 120 which upon reaction with trimethylsilylimine 121 yielded homoallylamine 122 (89% chemical yield 92% ee) as shown in Scheme 19. This H OH Ph O Me B(allyl)3 B Ph N Me HN Ts H Ts 119 N TMS 121 Ph Scheme 19 120 Ph NH2 122 (89% yield; 92% ee) represents the highest selectivity realized in the enantioselective allylation of various types of imines.34 Acyclic chiral 1,2-amino alcohols can be prepared by several fundamental routes (i) reduction of chiral amino acids; (ii) epoxide-opening reaction by the use of nitrogen nucleophiles such as amines and azides; (iii) selective ring opening reaction of 1,2-cyclic sulfates with a wide variety of amines and azides; (iv) stereoselective addition of nucleophiles to a-amino aldehydes and ketones and osmium-catalyzed asymmetric aminohydroxylation of olefins in the presence of chiral ligands.As described in section 2 methods for the synthesis of N-activated 2-aziridinemethanols are well established; however except for a few cases,13,27 there have been no systematic investigations toward a simple and effective method for synthesizing chiral N-protected 1,2-amino alcohols from readily available chiral 2-aziridinemethanols.Aza-Payne rearrangement of 2-aziridinemethanols followed by reacting with appropriate nucleophiles seems to meet this demand in terms of the stereoselectivity and efficiency. Chemical Society Reviews 1998 volume 27 152 The Payne rearrangement of a 2,3-epoxy alcohol to an isomeric 1,2-epoxy alcohol except for the recent protocol developed by Page Rayner and Sutherland,6 usually requires a basic aqueous medium a requirement that places a serious restriction on the types of nucleophilic agents which may be used in the subsequent epoxide-ring opening reaction in a onepot manner. We anticipated being able to synthesize stereochemically well defined functionalized enantiomerically pure N-protected 1,2-amino alcohols 126 in a stereo- and regio-selective sense starting from 2-aziridinemethanols such as 123 via anionic intermediates 124 and 125 by successive treatment with base and various nucleophilic reagents (epoxide ring-opening of the intermediate 125 at the less substituted carbon of the epoxide has been well documented).Thus whereas stereospecific synthesis of the anti-amino alcohol 126 could be expected starting from 2,3-trans-aziridino alcohol 123 syn-amino alcohol 128 could be stereospecifically synthesized from 2,3-cisaziridino alcohol 127 in a one-flask manner (Scheme 20). The rearrangement-epoxide ring opening reaction scenario does in fact lead to the stereochemically pure 1,2-amino alcohols.SO2R2 H N 1 O 3 R1 2 H 123 HN SO2R2 H Nu R1 H OH 126 SO2R2 H N H R1 2 3 1 OH 127 Initial experiments with 2-aziridinemethanol 14 revealed that KH was superior to either NaH or LDA as a base and KH was therefore used for all experiments. Reaction of 14 29 and 80 with KH (2 equiv.) in THF at 0 °C for 1 h was followed by the addition of Me2Cu(CN)Li2·2LiBr (5 equiv.) or BunCu(CN)Li2 in a one-pot manner and the mixture was then stirred for 1 h to yield the corresponding N-protected 1,2-amino alcohols in high yields as shown in Scheme 21. The presence of an excess of KH does not exert any influence on the organocopper reactions of intermediate epoxy aza-anions of type 125 in Scheme 20.It should be clearly noted that the use of MeLi·LiBr or MeMgBr instead of Me2Cu(CN)Li2·2LiBr did not result in clean transformations.35 The usefulness of this one-flask reaction would be enhanced if it could be successfully extended to other nucleophiles. Alkylthio and arylthio groups are important functional groups in various types of chemical transformations and as shown in Scheme 22 the reactions of 25 and 70 with KH followed by PhSH in a one-pot manner gave the diastereomerically pure phenylthio amino alcohols 132 and 133 in 92 and 89% yields respectively. Although similar treatment with KH and ButSH yielded only the corresponding tert-butylthio amino alcohols the overall yields of products were considerably lower (60–70%).35 N-Protected 1,2-amino alcohols bearing a nitrile group could also be prepared in a one-pot manner as shown in Scheme 23.The reaction of 70 and 135 with KH followed by the sequential SO2R2 :B H N O– R1 H H 124 SO2R2 –N :Nu i Nu R1 ii H3O+ O H H 125 i B HN SO2R2 H Nu ii Nu R1 H OH iii H3O+ 128 Scheme 20 Me OH OH 3 2 H i KH R Me H H 1 N ii Me2Cu(CN)Li2 Ts or Bun 2Cu(CN)Li2 14 H Ts—N H 17 R = Me (90%) R = Bun (88%) 129 Me OH OH H i KH R Me H H N ii Me2Cu(CN)Li2 H Ts—N Ts or Bun 2Cu(CN)Li2 29 H 18 R = Me (99%) R = Bun (89%) 130 Ph OH OH H i KH Me Ph H H N ii Me2Cu(CN)Li2 H Ts—N Ts H 80 (93%) 131 Scheme 21 All reactions were carried out with 2 mol equiv.of KH and 5 mol equiv. of organocopper reagents BnO OH OH H i KH SPh H BnO H N ii PhSH H Ts—N (92%) Ts H 25 132 OH Ph OH H SPh Ph i KH H H N H ii PhSH Ts—N Ts (89%) H 133 70 Scheme 22 All reactions were carried out with 2 mol equiv. of KH and 1.2 mol equiv. of PhSH OH Ph OH H CN Ph i ii iii H H N H (85%) Ts—N Ts H 134 70 BnO OH OH H i ii iii CN H BnO H H (95%) Ts—N H N Ts 135 (0.2 equiv.); iii Bu4 nNF 136 Scheme 23 Reagents i KH (2 equiv.); ii Me3SiCN (3 equiv.) Yb(CN)3 addition of Me3SiCN (3 equiv.) in the presence of Yb(CN)3 (0.2 equiv.) and Bu4NF (1.5 equiv.) yielded the nitriles 134 and 136 respectively in high yields after flash chromatographic purification.This reaction also appears to be quite general giving yields which are good to excellent.35 As shown in Scheme 24 synthesis of diamino alcohols 137 and 138 from aziridines 29 and 70 was also investigated. All attempts to use either dibenzylamine or lithium dibenzylamide to effect the epoxide ring opening of an anionic intermediate derived from 2-aziridinemethanol 29 were unsuccessful. Fortunately use of the higher order amide cuprate (Bn2N)2Cu(CN)Li2 overcame the problem and yielded diastereomerically pure diamino alcohols 137 and 138 in high yields from aziridinemethanols 29 and 70. In view of the synthetic utility of the above described reactions it was of interest to examine whether these trans- Me OH OH H NBn Me i KH ii [Bn2N]2Cu(CN)Li2 2 H H N H (91%) TsNH Ts 137 29 OH H Ph OH NBn Ph 2 i KH ii [Bn2N]2Cu(CN)Li2 H H H N Ts—N (91%) H Ts 138 70 OH OH H BnO Bn = benzyl Scheme 24 Reactions were carried out with 2 mol equiv.of KH and 5 mol equiv. of organocopper reagents formation products could be useful as synthetic intermediates. For example (2R,3S)-C18-dihydrosphingosine 140 can be readily synthesized in good overall yields from 2-aziridinemethanol 135 as shown in Scheme 25 via a sequence of reactions.35 i ii H H H N Ts–N (93%) Ts H 139 (CH2)13CH3 steps OH H HO H NH2·HCl BnO equiv.) 135 7 Concluding remarks (CH2)13CH3 153 140 Scheme 25 Reagents i KH (2 equiv.); ii [Me(CH2)13]2Cu(CN)Li2 (5 The aza-Payne rearrangement of activated 2-aziridinemethanols with bases such as KH and NaH in most common solvents such as THF toluene or a mixed solvent of THF–HMPA followed by quenching at low temperature gives the corresponding epoxy sulfonamides.Upon exposure of 2,3-epoxy amines to an equimolar mixture of ButOK–BunLi in a mixed solvent of THF and n-hexane at 278 °C the equilibrium lies exclusively toward the hydroxy aziridine forming direction. Although yields were not necessarily optimized an attractive one-pot regio- and stereo-selective synthetic route to chiral 1,2-amino alcohols from readily available 2-aziridinemethanols via the aza-Payne rearranged intermediates has been described. Isolation or purification of intermediates resulting from the aza-Payne rearrangement is not necessary.This methodology leads to a series of useful diastereomerically pure 1,2-amino alcohols which can be utilized as chiral auxiliaries for asymmetric synthesis. In addition the amino alcohols could be used for the synthesis of more complex molecules. The chemistry of the aza- Payne rearrangement reported herein increases our understanding of this relatively unexplored class of reactions. Many of the examples presented here have been developed with a view to large-scale laboratory synthesis in a simple manner. It is clearly evident even from the selective results described in this review that the aza-Payne rearrangement of 2-aziridinemethanols and 2,3-epoxy amines offer enormous potential for the synthesis of useful materials in enantiomerically pure form.The use of the aza-Payne reaction seems to be increasing and it would be apparent that researchers are Chemical Society Reviews 1998 volume 27 considering this strategy as a valuable alternative to the wellestablished routine but lengthy routes. 8 Acknowledgements None of our work described in this review could have been accomplished without the help of Professor Nobutaka Fujii (Kyoto University) Professor Yoshinori Yamamoto (Tohoku University) and many co-workers. The author thanks to Professor Timothy Gallagher (University of Bristol) for reading the manuscript and providing useful comments. This work was supported in part by Grant-in-Aid for Scientific Research (B) and (C) from the Ministry of Education Science Sports and Culture of Japan.9 References 1 W. H. G. Lake and S. Peat J. Chem. Soc. 1939 1069. 2 S. J. Angyal and P. T. Gilham J. Chem. Soc. 1957 3691. 3 J. G. Buchanan and H. Z. Sable in ‘Selective Organic Transformations’ ed. B. S. Thyagarajan Wiley New York 1972 vol. 2 p. 1; N. R. Williams in ‘Advances in Carbohydrate Chemistry and Biochemistry’ ed. R. S. Tipson and D. Horton Academic Press New York 1970 vol. 25 p. 109. 4 G. B. Payne J. Org. Chem. 1962 27 3819. 5 C. H. Behrens S. Y. Ko K. B. Sharpless and F. J. Walker J. Org. Chem. 1985 50 5687. 6 P. C. B. Page C. M. Rayner and I. O. Sutherland J. Chem. Soc. Perkin Trans. 1 1990 1375. 7 Y. Gao R. M. Hanson J. M. Klunder S. Y. Ko H.Masamune and K. B. Sharpless J. Am. Chem. Soc. 1987 109 5765. 8 T. Ibuka N. Mimura H. Aoyama M. Akaji H. Ohno Y. Miwa T. Taga K. Nakai H. Tamamura N. Fujii and Y. Yamamoto J. Org. Chem. 1997 62 999 and references cited therein. 9 Q. Liu A. P. Marchington N. Boden and C. M. Rayner J. Chem. Soc. Perkin Trans. 1 1997 511. 10 Q. Liu M. J. Simms N. Boden and C. M. Rayner J. Chem. Soc. Perkin Trans. 1 1994 1363. 11 C. M. Rayner Synlett 1997 11. 12 J. Moulines P. Charpentier J.-P. Bats A Nuhrich and A.-M. Lamidey Tetrahedron Lett. 1992 33 487. 13 R. S. Atkinson J. Fawcett D. R. Russell and P. J. Williams Tetrahedron Lett. 1995 36 3241 and references cited therein. 14 F. Latif A. Malik and W. Voelter Liebigs Ann. Chem. 1987 717. Chemical Society Reviews 1998 volume 27 154 15 R. Najime S. Pilard and M. Vaultier Tetrahedron Lett. 1992 33 5351. 16 H. Urabe Y. Aoyama and F. Sato Tetrahedron 1992 48 5639. 17 W. Kowollik G. Janairo and W. Voelter Liebigs Ann. Chem. 1988 427. 18 T. Ibuka K. Nakai H. Habashita N. Fujii F. Garrido A. Mann Y. Chounan and Y. Yamamoto Tetrahedron Lett. 1993 34 7421. 19 T. Ibuka K. Nakai H. Habashita Y. Hotta A. Otaka H. Tamamura N. Fujii N. Mimura Y. Miwa T. Taga Y. Chounan and Y. Yamamoto J. Org. Chem. 1995 60 2044. 20 N. Fujii K. Nakai H. Habashita Y. Hotta H. Tamamura A. Otaka and T. Ibuka Chem. Pharm. Bull. 1994 42 2241. 21 J. Åhman T. Jarevång and P. Somfai J. Org. Chem. 1996 61 8148 and references cited therein. 22 J. G. H. Willems F. J. Dommerholt J. B. Hammink A. M. Vaarhorst L. Thijs and B. Zwanenburg Tetrahedron Lett. 1995 36 603 and references cited therein. 23 G. E. Ham J. Org. Chem. 1964 29 3052. 24 C. H. Behrens and K. B. Sharpless J. Org. Chem. 1985 50 5696. 25 J. G. H. Willems M. C . Hersmis R. de Gelder J. M. M. Smits J. B. Hammink F. J. Dommerholt L. Thijs and B. Zwanenburg J. Chem. Soc. Perkin Trans. 1 1997 963. 26 C. Gennari A. Vulpetti and G. Pain Tetrahedron 1997 53 5909. 27 F. A. Davis P. Zhou and G. V. Reddy J. Org. Chem. 1994 59 28 D. A. Evans M. M. Faul M. T. Bilodeau B. A. Anderson and 29 K. B. Hansen N. S. Finney and E. N. Jacobsen Angew. Chem. Int. Ed. 30 For a recent review see H. M. I. Osborn and J. Sweeney Tetra- 31 M. Vaultier N. Knouzi and R. Carri�e Tetrahedron Lett. 1983 24 32 M. Schlosser in ‘Organometallics in Synthesis’ ed. M. Schlosser John 33 U. M. Lindstrom R. Franckowiak N. Pinault and P. Somfai 34 S. Itsuno K. Watanabe K. Ito A. A. El-Shehawy and A. A. Sarhan 35 T. Ibuka K. Nakai M. Akaji H. Tamamura N. Fujii and Y. Yamamoto 3243. D. M. Barnes J. Am. Chem. Soc. 1993 115 5328. Engl. 1995 34 676. hedron Asymmetry 1997 8 1693 and references cited therein. 763. Wiley and Sons New York 1994 pp. 1–166. Tetrahedron Lett. 1997 38 2027 and references cited therein. Angew. Chem. Int. Ed. Engl. 1997 36 109. Tetrahedron 1996 52 11 739 and references cited therein. Paper 7/05976K Received 14th August 1997 Accepted 21st October 19

ISSN:0306-0012    

DOI:10.1039/a827145z

出版商:RSC    

年代:1998

数据来源: RSC

摘要:

Carbocycles from carbohydrates via free radical cyclizations new synthetic approaches to glycomimetics Angeles Mart�ýnez-Graua and Jos�e Marco-Contelles*,b a anica Facultad de Qu Departamento de Qu�ýmica Org� �ýmica Ciudad Universitaria s/n 28040-Madrid Spain b anica Juan de la Cierva 3 28006-Madrid Spain Instituto de Qu�ýmica Org� Free radical cyclization of enantiomerically pure acyclic presursors derived from carbohydrates is an excellent method for the synthesis of complex densely functionalized chiral carbohydrate mimics (‘glycomimetics’). The extent of the acyclic diastereoselection can be modulated and is closely associated with the structure rigidity and conformational aspects of the radical precursors. General models for a rationale of the stereochemical outcome of the cyclizations are shown.1 Introduction Carbohydrates are readily available and inexpensive building blocks for the synthesis of natural products.1 In the last decade particular emphasis has been devoted to the preparation of polyfunctionalized carbocycles from sugars.2 Carbocyclization based on free radical chemistry has been recognized as an efficient tool because radical cyclization methods tolerate high levels of substrate functionalization.3 The general process includes monosaccharide protection opening functionalization and finally cyclization to a carbocyclic ring of five or six carbons. These efforts have culminated in the design of some new free radical based methodologies for the synthesis of mimetics of natural sugars (‘glycomimetics’) such as aminocyclopentitols inositols branched chain cyclitols or carbasugars.Glycomimetics play critical roles in biological systems and by Angeles Mart�ýnez-Grau was born in Madrid (Spain) in 1966. She received her BSc degree in chemistry from the Universidad Complutense of Madrid in 1990. She obtained her MSc at Instituto de Qu�ýmica Org�anica of Madrid (CSIC) in 1990 and her PhD in organic chemisry at the Universidad Complutense in 1994. She was a postdoctoral fellow with Professor Dennis P. Curran at the University of Pittsburgh (USA) from 1995 to 1996. Her research has focussed on free radical chemistry and stereoselective synthesis. In 1996 she was awarded a Spanish Royal Society of Chemistry Prize for young scientists.She joined the faculty in the Department of Organic Chemistry at the Universidad Complutense of Madrid in 1996. Angeles Mart�ýnez-Grau • RO RO I a O O c controlling cell–cell interactions or interfering with carbohydrate metabolic processes are potential candidates for new pharmaceutical drugs.4 In the last nine years our group has been actively working in this area. In this review we present the most relevant results for the synthesis of aminocyclopentitols and cyclitols by free radical cyclization of acyclic monosaccharide derivatives. Particular emphasis is placed on the stereochemical aspects of these reactions. Most of the conversions are straightforward examples that illustrate the utility scope and generality of this strategy for the synthesis of enantiomerically pure carbocycles.2 Aminocyclopentitols via cyclization of d-functionalized O-alkyl oxime ethers derived from carbohydrates The aminocyclopentitol nucleus is present in a wide variety of natural products. The synthesis of these polyhydroxylated cyclopentylamines has been an area of growing interest due to the discovery of their potent biological properties. For example allosamidin (1) trehazolin (3) mannostatins (4) and Merrel Dow’s cyclopentylamine (5) have been found to have powerful and specific inhibitory activity against glycosidases. As a consequence they have attracted the interest of organic chemists and many synthetic approaches have been reported.5 Another important group of aminocyclopentitols the 4-amino- 1,2,3-cyclopentanetriols are also versatile and valuable inter- Born in 1956 Jos�e Marco-Contelles received his BSc in chemistry from the Universidad Complutense of Madrid (Spain).He obained his PhD at the Instituto de Qu�ýmica Org�anica of Madrid (CSIC) and did his postdoctoral studies at Gif-sur-Yvette (Institute de Substances Naturelles CNRS France) in H.-P. Husson’s laboratory (1984–1985). In 1986 he was a postdoctoral fellow with Professor W. Oppolzer at Geneva University (Switzerland). From 1988 to 1989 he was an Associate Researcher at Duke University (USA) working with Professor B. Fraser-Reid. In 1985 he was appointed as Associate Researcher at the Instituto de Qu�ýmica Org�anica of Madrid (CSIC) and promoted to Scientific Researcher in 1992.His current research areas include the development of new synthetic methodologies in free radical chemistry carbohydrate chemistry asymmetric synthesis and synthesis of new heterocycles. He is the author of more than 100 papers and reviews. Jos�e Marco-Contelles 155 Chemical Society Reviews 1998 volume 27 mediates for the synthesis of cyclopentane-type glycosidase inhibitors and carbocyclic nucleosides.6 The 5-exo radical cyclization reaction of acyclic carbohydrate derivatives has been applied to the synthesis of a broad range of polyhydroxylated cyclopentanes using different types of radical acceptors olefins and enol ethers,7 unsaturated esters,8 alkynes,9 imines,10 oxime ethers11 and hydrazones.12 In this section we present a range of free radical cyclizations of acyclic d-functionalized O-alkyl oxime ethers derived from carbohydrates to obtain different aminocyclopentitols.OH RO O CH3 N N HO CH3 (–)-Allosamidin (1) OH OH O O HO R = O NHAc ) NHAc OH OH (–)-Allosamizoline (2 R = H) NH2 X HO OH HO Mannostatin A (4a X = SCH3) Mannostatin B [4b X = S(O)CH3] 2.1 Bu3SnH-mediated 5-exo radical cyclizations The ability of O-alkylaldoximes to act as intramolecular radical traps is well known. Bartlett was the first to synthesize aminocyclopentitols by cyclization of 5-oximinoalkyl radicals derived from d-glucose (Scheme 1).11 Four stereoisomers can arise in the cyclization of radical precursor 6; however only carbocycles 7 and 8 were obtained in a 62 38 ratio respectively.The 1,5-cis product is the major one and this result can be explained by assuming that the intermediate radical preferentially adopts a chair-like conformation with the substituents at C2 C3 and C4 in pseudoequatorial positions.13 NOBn S OBn PhO O Bu3SnH AIBN OBn OBn BnO 6 Scheme 1 Following a similar strategy Simpkins designed a short enantiospecific route to allosamizoline (2) the aminocy- Chemical Society Reviews 1998 volume 27 156 OH OH HO OH HO O N OH N O HO H OH Trehazolin (3) OH NHCH HO 3 OH HO Merrel Dow¢s cyclopentylamine (5) NHOBn 1 5 OBn BnO OBn BnO 7 (62:38) + NHOBn 1 5 OBn BnO OBn BnO 8 NOBn S NHCbz Im O OAc OAc AcO 9 2 X R1O NOR O O 12a-f Br O NOBn Ph O OTBS 15 clopentitol aglycon of allosamidin (1) starting from d-glucosamine (Scheme 2).14 The cyclization of radical precursor 9 showed poor diastereoselectivity; three compounds were obtained the minor 10 and the major 11 isolated as a mixture of epimers in C1.In our group we carried out a new synthesis of enantiomerically pure 4-amino-1,2,3-cyclopentanetriols15 via free radical cyclization of a series of oxime ethers 12a–f (Scheme 3) and 15 (Scheme 4); these compounds were easily prepared from d-ribono-g-lactone. All precursors were obtained and used as inseparable mixtures of E and Z oxime isomers (7 3 ratio respectively) and were submitted together to typical conditions for radical cyclization promoted by tributyltin hydride or samarium diiodide.+ NHOBn 1 11 O O O O NHOBn OAc NHCbz OAc AcO 10 OAc NHCbz OAc AcO Scheme 2 4 NHOR R1O Bu3SnH or SmI2 O 13a-f + NHOR R1O O 14a-f Scheme 3 NHOBn Bu3SnH or SmI2 Ph O OTBS 16 + NHOBn Ph O OTBS 17 Scheme 4 From precursors 12a–f (Scheme 3) using tributyltin hydride all the cyclizations showed good yields and high sterethe formation of the exo products 13a–f. When samarium diiodide was used the yields were lower due to several competititve processes epoxide ring formation and/or elimination; however compounds 13a–f were the only diastereomers obtained.15 Table 1 Free radical cyclization of precursors 12 using Bu3SnH Total yield (%) 13/14 R1 12 R X H ButMe2Si 75 53 52 58 71 85 100/0 100/0 100/0 89/11 80/20 100/0 Bn Bn Bn Bn Me Bn Ac Bz Bz H Br Br Br Br Br I a bc def The cyclization of precursor 15 (Scheme 4) using the tin hydride method provided a mixture of 16 and 17 in a 1.8 1.0 ratio and 80% total yield.The samarium diiodide ring closure gave the major product 16 in 40% yield with traces of 17.15 The stereochemical outcome of these free radical cyclizations is consistent with the transition state model proposed by Beckwith,13 in which the radical species is in a chair-like conformation having most of the substituents in the preferred pseudo-equatorial positions.This is in agreement with the results reported by Wilcox for the cylization of analogous a,bunsaturated esters.8 In Scheme 5 hypothetical lowest energy transition states for the intermediate radicals are presented. In conformation B the unfavourable 1,3-diaxial interaction between the substituents at C2 and C4 (radical numbering) destabilizes the transition state leading to the endo product. We have not found a satisfactory explanation for the formation of small amounts of the endo product during the cyclization of precursors 12d–e (Table 1) with a benzoyl group at R1 (compare with the results obtained in the cyclization of precursors 12a–c,f with different protecting groups at this position).The higher diastereoselectivity observed in samarium-mediated carbocyclization compared with tributyltin hydride cyclization is difficult to rationalize at this moment. We can hypothesize that a chelated samarium–ligand transition state12 should impose some limits to the conformational freedom in the reactive species leading to higher stereochemical control. OR 2 N • • R1O 4 O N OR OR1 O O O B A 1 4 NHOR R1O NHOR R1O 3 2 O O O O 14a-f 13a-f Scheme 5 In summary the free radical cyclization of d-functionalized O-alkyl oxime ethers derived from d-ribose is a good method for the synthesis of (1R,2R,3S,4R)-4-amino-1,2,3-cyclopentanetriol derivatives in terms of chemical yield and acyclic diastereoselection.Major trans (C1–C4 and C3–C4) isomers are formed (Scheme 5). Comparing both cyclizations methods the tin-mediated cyclization of 5-oximinoalkyl radicals derived from monosaccharides is clearly superior to the corresponding SmI2-mediated reaction in terms of chemical yield and experimental manipulation. The use of dithioketals as a radical precursor in this type of cyclization has also been studied. Roberts was the first to use as a key step to prepare new mannostatin analogues a free radical carbocyclization of an a-sulfenyl radical on oxime ethers starting from d-allose (Scheme 6).16 The cyclization of radical precursor 18 occurs in high yield (80%) but the stereoselectivity was moderate (3 1) and unfortunately both isomers 19 and 20 were obtained as an inseparable mixture.MeON H OAc SEt Bu3SnH SEt AIBN BnO 18 Scheme 6 In parallel work we reported the synthesis of new mannostatin A analogues by free radical cyclization of precursor 21 prepared from d-ribose via a shorter route (Scheme 7).17 The yield was also high (80%) and although the stereoselectivity was only slightly better (4 1) the major compound 22 was 20 NHOBn BnO Bu3SnH OAc isolated in almost pure form. SEt AIBN BnO NOBn SEt 21 Scheme 7 OBn N BnO The cis stereochemistry at the new stereocenters (C1 and C5) in the major compound can be explained by assuming a chairlike conformation in the transition state of the 5-exo-trig ring closure (Scheme 8) with the oxime ether and most of the substituents in the preferred pseudo-equatorial positions (conformation C) to avoid 1,3-diaxial interactions in other possible conformers.23 NHOBn 1 2 BnO C Scheme 8 OBn SEt • OBn Chemical Society Reviews 1998 volume 27 NHOMe SEt AcO OAc BnO (3:1) 19 + NHOMe SEt AcO OAc BnO SEt BnO OBn BnO (4:1) 22 + NHOBn SEt BnO OBn BnO 5 4 SEt BnO 3 OBn BnO 22 157 R3SnH OR R 1 2 3 4 O O 24 O O R Et3B Scheme 9 R1 H Ph H Bu Ac Ph TBDMS Ph Scheme 10 O Entry D Table 2 Free radical cyclization of precursor 24 BnOHN H BnOHN O 2.2 R3SnH-mediated 5-exo cyclization of d-alkyne tethered O-alkyl oxime ethers The intramolecular coupling of oxime ethers tethered to terminal alkynes to prepare acyclic sugar derivatives has been carried out using a hydrostannylation reaction.18 Derived from d-mannose the radical precursor 24 (Scheme 9) was cyclized using tributyltin or triphenyltin hydride plus Et3B to give the vinyl stannanes 25 as a mixture of the separable Z/E isomers (Table 2).Each geometric isomer was diastereomerically pure at the new stereocenter (C4) with the absolute configuration shown in Scheme 9. The synthesis and cyclization of a similar radical precursor prepared from d-ribose leads to enantiomers of compounds 25Z/E in good yield and complete diastereoselection at the newly formed stereocenter.The diastereoselectivity observed in the cyclization can be explained according to Beckwith’s guidelines,13 assuming that in the early transition state the favoured intermediate vinyl radical is in a chair-like conformation D with most of the substituents in the preferred pseudoequatorial orientation (Scheme 10). As in the cyclization of precursors 12 (Scheme 3) major isomers having a trans relationship between C1/C4 and C3/C4 are obtained. 25E 25Z (%) 25E (%) SnR1 3 1 2 O 25 BnON H 158 SnR3 • N OBn 2.3 SmI2-mediated intramolecular reductive coupling of d-carbonyl tethered O-alkyl oxime ethers In 1989 Enholm was the first to describe the samariummediated intramolecular cyclization of aldehydes tethered to Chemical Society Reviews 1998 volume 27 H SnR1 3 4 OR BnOHN O O 25Z + H OR O O 3 69 78 (4.5 1 ratio) — 14 70 45 OR R1 3Sn 4 3 a,b-unsaturated esters in substrates derived from carbohydrates.19 Our group has recently developed the samarium-mediated intramolecular reductive coupling of carbonyl-tethered O-alkyl oxime ethers to provide a new and synthetically useful method for the construction of aminocyclopentitols.15 This procedure was applied to the synthesis of trehazolamine analogous 26 (Scheme 11) by reductive coupling of d-carbonyl tethered O-alkyl oxime ethers derived from d-glucose. Reaction of precursor 27 using samarium iodide gave only isomer 28 in 80% yield. Simultaneously Naito also analyzed this protocol using Bu3SnH as a reagent for carbocyclization.20 Reaction of 27 (Scheme 11) with tributyltin hydride gave a 1.0 1.4 mixture of 28 and 29 respectively in 68% overall yield.This reaction using samarium diiodide was also carried out with precursors obtained from d-mannose and d-galactose.15 Ketone 30 derived from d-mannose afforded a mixture of three aminocyclopentitols 31–33 (Scheme 11) in a ratio 15 3 1 respectively. The reductive coupling of precursor 34 obtained from d-galactose gave only compound 35. In all cases the cyclization gave major compounds showing a trans relationship between the hydroxy at C1 and the benzyloxyamino group at C5 and between C4 and C5. There is also a trans relationship between the hydroxy at C1 and the benzyloxy group at C2 (except for the cyclization of the d-glucose derivative 27 that led to isomer 28).The stereochemistry obtained was independent of the geometry of the starting oxime ether. Surprisingly no additives such as HMPA were necessary. These results can be rationalized in terms of a ninemembered ring chelate12 as shown in Scheme 12. The anion radical species adopt chair-like conformations in which the O-alkyl oxime ether at C5 (radical numbering) is in the preferred pseudoequatorial position. In order to explore the scope of the method related aldehydes were also studied.15 To this end alcohol 36 prepared from d-arabinose was submitted to a one-pot sequence of Swern oxidation and reductive coupling promoted by SmI2 (Scheme 13).A mixture of 37a and 37b was obtained in a 8 1 ratio respectively. Following the same procedure appropriate and conveniently functionalized alcohols derived from d-ribose and d-xylose were also cyclized in good yield and from excellent to moderate diastereoselectivity.15 In the SmI2-promoted cyclization of d-aldehyde O-alkyl oxime ethers we have observed a trans relationship between C1/C2 and C4/C5 in the major isomers. Collectively the experiments above suggest that the samarium diiodide mediated 5-exo intramolecular cyclization of d-carbonyl tethered O-alkyl oxime ethers derived from sugars is a good method for the synthesis of aminocyclopentitols in enantiomerically pure form. 2- 2.4 SmI2-mediated intramolecular reductive coupling of a,b-unsaturated esters tethered O-alkyl oxime ethers As a complement of the well known SmI2-mediated reductive coupling of sugar-derived a,b-unsaturated esters tethered to carbonyl groups,19 we have also described the first SmI promoted intramolecular cyclization of unsaturated esters to O-alkyl oxime ethers (Scheme 14).15 The cyclization of E-isomers 38 provided compound 39 in 52% yield as a single isomer.3 Cyclitols from sugars via free radical cyclization Cyclitols and aminocyclitols are polyhydroxylated cyclohexanes with important biological activities.4 Different methodologies have been developed for the synthesis of these compounds. 21 Sugars have been commonly used as chiral starting materials. However among the numerous methods reported for the conversion of carbohydrates into cyclitols free radical cyclization is not so frequently used when a cyclohexane is the target ring.OBn N BnO BnO 27 BnO O OBn 30 OBn BnO N BnO O 1 • 5 OBn • BnO 34 BnO BnO BnO 1 • 5 H H SmI2•(HMPA) n OBn N 30 BnO OBn SmI2•(HMPA) n OBn OBn SmI2•(HMPA) n 1 5 O Scheme 12 3.1 Bu3SnH-mediated 6-endo cyclization Hex-5-ynyl radicals cyclize preferentially via the kinetically favoured 5-exo pathway. Nevertheless the regioselectivity can be changed by structural factors.22 This feature has been used in a novel strategy for the synthesis of cyclitols.23 We thought that in the cyclization of 1-iodoalk-5-ynes of type 40 (Scheme 15) the 6-endo-dig ring closure should be preferred because the trans located 1,3-dioxolane moiety prevents the 5-exo-dig pathway leading to thermodynamically disfavoured trans fused cyclopentanes.The cyclohexenes 41 were obtained in good yield; the diastereomeric mixture of 41a was oxidized to cyclohexanone 42 a highly functionalized chiral intermediate for the synthesis of cyclitols.24 O NOR BnO BnO OBn OBn 27 HO OH NH HO 2 OH HO O NOBn BnO OBn OBn O NOBn BnO OBn 26 Trehazolamine SmI2 THF OBn 34 28 31 35 BnO OH 1 SmI2 THF NHOR BnO 2 5 ref. 15 4 OBn BnO BnO OH BnO BnO 29 BnO BnO OH OH X 1 NHOBn BnO 5 2 Y 4 OBn OBn BnO BnO 32 X = H; Y = NHOBn 33 Y = H; X = NHOBn BnO 1 NHOBn BnO 5 2 OBn BnO OH OH 1 5 2 NHOBn a) Swern oxidation O O 4 b) SmI2 ButOH 3 O O OTBS 37a 36 (8:1) + OH NHOBn O O OTBS 37b Scheme 13 CO2Et 2Et NHOBn NOBn SmI2 O O THF/HMPA O Ph O Ph OTBS 39 (52%) 38 Scheme 14 28 NHOR OBn + OH 4 35 NOBn OTBS CO OTBS Chemical Society Reviews 1998 volume 27 31 SmI2 THF Scheme 11 3.2 Bu3SnH-mediated 6-exo cyclization Most of the free radical cyclizations of acyclic carbohydrate derivatives to obtain carbocycles use hex-5-enyl radicals whereas the 6-exo closure to give cyclohexane rings has been used less frequently because the cyclization of hept-6-enyl radicals is about 40 times slower than the cyclization of the hexenyl radical.Therefore reactions competing with the 159 • RO RO RO O b I a O O O O O O 40a R = H b R = Ac 41a R = H 50 % b R = Ac 66% Scheme 15 cyclization (e.g. reduction of the initial radical) become important. Another potential problem of the heptenyl radical is due to a 1,5-hydrogen transfer that leads to allyl radicals. Nevertheless the 6-exo cyclization is a well known procedure in the synthesis of branched chain sugars where the cyclization of hept-6-enyl radicals is accelerated by the introduction of electron-withdrawing groups in the radical acceptor.25 Crich obtained a 1 1 mixture of the cyclohexanones 44 and 45 (Scheme 16) in 90% yield after cyclization of an acyclic acyl radical.26 O SPh O SPh PhSe OTBS TBSO OTBS TBSO 44 43 + (1:1) O SPh OTBS TBSO 45 Scheme 16 Redlich carried out a very interesting study on the cyclization of 7-deoxy-7-iodohept-1-enitols of the allo- manno- galacto- Br R1O Bu3SnH OR N R2O AIBN OR2 R1O Br Bu3SnH OBn AcO N 2 AIBN O Br Bu3SnH OBn AcO N 2 AIBN O 52 46a-e AcO 3 O 49 AcO 3 O Chemical Society Reviews 1998 volume 27 160 O O and gulo-series using tributyltin hydride.27 The cyclization of radical precursors provided carba-6-deoxyhexoses in very good yield.The authors studied the effect of varying the protecting groups adjacent to the reaction centers and the influence of the stereochemistry. They rationalized the observed selectivities in terms of chair- and boat-like transition states depending on the threo or erythro disposition of the substituents contiguous to the reaction centers.42 Of particular interest has been the application of the 6-exo free radical cyclization to the synthesis of aminocyclitols. In our group we carried out the first synthesis of 5-amino-1,2,3,4-cyclohexanetetrols by 6-exo cyclization of 6-bromo-6-deoxyhexose derivatives using oxime ethers as radicals acceptors (Scheme 17).28 All the precursors were obtained as inseparable mixtures of E and Z isomers in a 70 30 ratio. The radical is generated at an appropriate distance from the double bond to give 6-exo ring closure. The precurosrs 46a–e prepared from d-glucose with different protecting groups (Table 3) and the d-manno 49 and d-gulo 52 derivatives were cyclized in moderate yield and good diastereoselectivity.The incorporation of an isopropylidenedioxy at C2 and C3 in radical precursors 49 and 52 dramatically increased the diastereoselection (Scheme 17) giving almost a single isomer at the newly formed stereocenter (C5). Table 3 Free radical cyclization of precursors 46 46 ab cd e The stereochemical outcome of the cyclizations could be justified by assuming that the major products arose via transition states in which the radical species adopts a preferred chair-like conformation with most of the substituents occupying pseudoequatorial positions to minimize unfavourable 1,3-diaxial steric interactions.28 In precursors 46 (Scheme 18) the steric interaction between the substituent at C5 and the oxime R1O NHOR R2O OR2 R1O 5 NHOBn AcO O O 97:3 (50%) 5 NHOBn AcO 47a-e AcO AcO O O 53 (50%) 50 Scheme 17 Yield (%) R2 47/48 R1 R 52 55 50 40 42 83/17 75/25 73/27 80/20 78/22 Bn Bn Me Bn Bn Ac Bz Bz H Ac Ac Bz Bz Bn Bn R1O NHOR R2O + OR2 R1O 48a-e AcO NHOBn AcO + O O 51 ether at C1 (radical numbering) displaces the equilibrium to conformer E leading predominantly to compound 47.Nevertheless this steric repulsion is not critical because the results of substrates 49 and 52 point to the significance of the gauche interaction between the C2 substituent (isopropylidenedioxy group) and the oxime ether at C1.5 • O O N OBn 1 O O H E R1O NHOR R2O R1O OR2 47 Scheme 18 This work has been extended to explore the 6-exo cyclization of several 6-bromo-6-deoxy-d-glucose derivatives by changing the nature of the radical trap.28 As expected the d-gluco derivative 54a with two isopropylidenedioxy groups at C2/C3 and C4/C5 afforded after cyclization only isomer 55a in 82% diastereomeric excess (Scheme 19). The reactions of enol ether 54b and a,b-unsaturated ester 54c with tributyltin hydride gave the major carbocycles 55b–c in high diastereomeric excess (84 and 86% respectively). This improvement in the stereoselectivity compared with that observed for the analogous precursors 46a–e confirms the influence of structural factors on the stereochemical outcome of the reaction.Br O O O X Bu3SnH AlBN O O 54a X = CH NOBn b X = CH CHOMe c X = CH CHCO 55a Y = NHOBn (d.e. 82%; 75%) b Y = CH2OMe (d.e. 84%; 60%) c Y = CH2CO2Me (d.e. 86%; 80%) 2Me Scheme 19 After our initial reports on the synthesis of branched chain cyclitols,28 several groups have successfully applied this strategy to the synthesis of some natural products.29,30 3.3 SmI2-mediated intramolecular reductive coupling of e-carbonyl tethered O-alkyl oxime ethers Continuing our interest in the synthesis of aminocyclitols we tried the intramolecular reductive coupling of a sugar-derived oxime ether e-functionalized with an aldehyde.15 In order to favor the cyclization we selected the conformationally strained d-glucose derivative 56 (Scheme 20).Dess–Martin oxidation gave the intermediate aldehyde that was submitted to reductive coupling promoted by samarium diiodide yielding the isomers 57a–c. Total yield was good but the diastereoselection was very poor. 4 Conclusions In summary we have shown that the 5-exo 6-endo and 6-exo free radical carbocyclization of acyclic conveniently functionalized radical precursors derived from sugars is a suitable method for the synthesis of polyhydroxylated carbocycles in 5 O O H O O 1 N OBn F R1O NHOR R2O OR2 R1O 48 O Y O O O OH O OH NHOBn O O NOBn O O O O 56 57a (22%) + O O OH OH NHOBn O NHOBn O O O O O 57b (24%) + Scheme 20 57c (20%) enantiomerically pure form.The type of radical trap allows us to manipulate the type of carbocycle to be prepared. The appropriate selection of radical precursor and the nature of the functional groups also determines the qualitative and quantitative course of the acyclic diastereoselection in the formation of new stereocenters. 5 References 1 P. Collins and R. R. Ferrier in Monosaccharides Wiley and Sons 2 T. V. RajanBabu in Preparative Carbohydrate Chemistry ed. S. 3 B. Giese in Radicals in Organic Synthesis Formation of Carbon- Chichester 1995. Hanessian Marcel Dekker New York 1997 p. 545. Carbon Bond Pergamon Oxford 1986. 4 A. D. Elbein Ann. Rev.Biochem. 1987 56 497. 5 See for instance B. M. Trost and D. L. V. Vranken J. Am. Chem. Soc. 1993 115 444. 6 A. D. Borthwick and K. Biggadike Tetrahedron 1992 48 571. 7 T. V. RajanBabu Acc. Chem. Res. 1991 24 139. 8 C. S. Wilcox and L. M. Thomasco J. Org. Chem. 1985 17 546. 9 J. C. Malanda and A. Doutheau J. Carbohydr. Chem. 1993 12 999. 10 M. J. Tomaszuvski J. Warkentin and N. H. Werstink Aust. J. Chem. 1995 48 291. 11 P. A. Bartlett K. L. McLaren and P. C. Ting J. Am. Chem. Soc. 1988 110 1633. 12 C. F. Sturino and A. G. Fallis J. Am. Chem. Soc. 1994 116 7447. 13 A. L. J. Beckwith Tetrahedron 1981 37 3073. 14 N. S. Simpkins S. Stokes and A. J. Whittle J. Chem. Soc. Perkin Trans. 1 1992 2471. 15 J. Marco-Contelles P. Gallego M. Rodr�ýguez-Fern�andez N. Khiar C. Destabel M. Bernab�e A. Mart�ýnez-Grau and J. L. Chiara J. Org. Chem. 1997 62 7397. 16 A. H. Ingall P. R. Moore and S. M. Roberts Tetrahedron Asymmetry 1994 5 2155. 17 J. Marco-Contelles C. Destabel and P. Gallego J. Carbohydr. Chem. 1995 14 1343. 18 J. Marco-Contelles C. Destabel P. Gallego J. L. Chiara and M. Bernab�e J. Org. Chem. 1996 61 1354. 19 E. J. Enholm and A. Trivellas J. Am. Chem. Soc. 1989 111 6463. 20 T. Kiguchi K. Tajiri I. Ninomiya T. Naito and H. Hiramatsu 21 P. Vogel D. Fatton F. Gasperini and C. Le Brian Synlett 1990 173 22 U. Albrecht R. Wartchow and H. M. R. Hoffmann Angew. Chem. Int. 23 J. Marco-Contelles M. Bernab�e D. Ayala and B. S�anchez J. Org. 24 X. Deruyttere L. Dumortier J. Van der Eycken and M. Vandewalle 25 J. Marco-Contelles A. Mart�ýnez-Grau M. Mart�ýnez-Ripoll F. H. Cano Tetrahedron Lett. 1995 36 253. and references cited therein. Ed. Engl. 1992 31 910. Chem. 1994 59 1234. Synlett 1992 51. and C. Foces-Foces J. Org. Chem. 1992 57 403. Chemical Society Reviews 1998 volume 27 161 26 D. Batty D. Crich and S. M. Fortt J. Chem. Soc. Chem. Commun. 1989 1366. 27 H. Redlich W. Sudau A. K. Szardenings and R. Vollerthun Carbohydr. Res. 1992 226 57. 28 J. L. Marco-Contelles C. Pozuelo M. L. Jimeno L. Mart�ýnez and A. Mart�ýnez-Grau J. Org. Chem. 1992 57 2625. Chemical Society Reviews 1998 volume 27 162 29 G. E. Keck S. F. McHardy and J. A. Murry J. Am. Chem. Soc. 1995 117 7289. 30 R. Clauss and R. Hunter J. Chem. Soc. Perkin Trans. 1 1997 71. Received 25th July 1997 Acce

ISSN:0306-0012    

DOI:10.1039/a827155z

出版商:RSC    

年代:1998

数据来源: RSC

摘要:

Non-conventional hydrogen bonds Ibon Alkorta Isabel Rozas and Jos�e Elguero Instituto de Qu�ýmica M�edica CSIC Juan de la Cierva 3 E-28006 Madrid Spain. E-mail ibon@pinar1.csic.es; rozas@pinar1.csic.es & jelguero@pinar1.csic.es Hydrogen bonds (HBs) are the most important ‘weak’ interactions encountered in solid liquid and gas phases. The HB can be defined as an attractive interaction between two molecular moieties in which at least one of them contains a hydrogen atom that plays a fundamental role. Classical HBs correspond to those formed by two heteroatoms A and B with a hydrogen atom bonded to one of them and located approximately in between (A–H···B). Recently knowledge of the number of functional groups which act as hydrogen bond donors or acceptors has increased considerably and most of these new groups are discussed.1 Introduction Hydrogen bonds (HBs) are the most important ‘weak’ interactions encountered in solid liquid and gas phases. They define the crystal packing of many organic and organometallic molecules the 3D structure of biological macromolecules as well as modulate the reactivity of different groups within a molecule. Hydrogen bonds (HBs) can be defined as an attractive interaction between two molecular moieties (two molecules or two parts of the same molecule) in which at least one of them contains a hydrogen atom that plays a fundamental role in the interaction. In the rank of interactions among atoms the HB falls between chemical bonds (as covalent bonds) and nonbonding interactions such as van der Waals interactions.In general a HB is characterized by (i) a weak to medium interaction energy;1 (ii) a considerable interpenetration of the Ibon Alkorta was born in Eibar (Basque Country) in 1963 obtained his PhD degree from the Universidad Aut�onoma de Madrid in 1990 and was a post-doctoral fellow with Dr Hugo Villar (Molecular Research Institute Palo Alto California 1991–1993). He presently works at the Instituto de Qu�ýmica M�edica as Colaborador Cient�ýfico and his current interests include the theoretical study of solvation and hydrogen bonds. Isabel Rozas was born in Madrid in 1959 obtained her PhD degree from the Universidad Complutense and was a post- Isabel Rozas Ibon Alkorta isolated electronic clouds of the two moieties involved; (iii) a certain electron transfer between the two moieties and (iv) a preferred geometry.2–4 Hibbert and Emsley1 define three kinds of HBs depending on the interaction energy values obtained.Thus HBs with energies between 22.4 and 212 kcal mol21 (1 cal = 4.184 J) are defined as weak HBs those with energies between 212 and 224 kcal mol21 are defined as strong HBs and those with energies more negative than 224 kcal mol21 are considered very strong HBs. The classical HBs correspond to those formed by two heteroatoms A and B with a hydrogen atom bonded to one of them and located approximately in between (A–H···B). In general there has been considered an electrostatic attraction between the positive end of the bond dipole of A–H and the centre of negative charge on B (generally a lone pair of electrons).Usually the A–H moiety is defined either as an ‘electron acceptor’ or as a ‘hydrogen bond donor (HBD)’ and the B moiety as an ‘electron donor’ or a ‘hydrogen bond acceptor (HBA)’. In this review we will use the hydrogen bond donor–acceptor nomenclature5,6 except in Section 8. The previously mentioned classical requirements for HBDs and HBAs mean that both moieties come mainly from the same few groups of the periodic table groups 15 16 and 17. Therefore most published works of HBs are the type O–H···B or N–H···B in which the HB acceptor ‘B’ posses N O or F lone pairs responsible for the HB formation. These classical HBs have been generalized in other directions such as (i) HBs with unconventional H donors such as C–H (ii) HBs with unconvendoctoral fellow with Professor Paul G.Mezey (University of Saskatchewan Canada 1989–1991) and Professor Donald F. Weaver (Queen’s University Canada 1993–1994). She presently works at the Instituto de Qu�ýmica M�edica and her current research interests include the study of hydrogen bonds and the design of new bradykinin antagonists. Jos�e Elguero was born in Madrid in 1934 and received his PhD degree from the University of Montpellier in 1961. He has worked in both the French CNRS and the Spanish CSIC where he is now Senior Research Scientist. He has been awarded the Gold Medal of the Spanish Royal Society of Chemistry the Schutzenberg Solvay and Ram�on-y-Cajal Prizes.His specialities are heterocyclic chemistry and physical organic chemistry fields in which he has published nearly 800 papers. Jos�e Elguero 163 Chemical Society Reviews 1998 volume 27 tional H acceptors as p-bonded functional groups halogens or C atoms and (iii) dihydrogen bonds A–H···H–B. Research over the years has widened the knowledge of groups acting as HBDs or HBAs our understanding of which is reported in Table 1. The present review will not include ‘classical’ HBs (A and B both being heteroatoms) nor the ‘nonconventional’ case when A is a carbon atom (C–H as an HBD) because they have been reviewed in detail elsewhere.7–9 From the point of view of the interaction energy the HBs considered in the present review are within those considered weak by Hibbert and Emsley;1 however for a better understanding of these unconventional HBs we will estabish a new classification HBs with interaction energies until ý25ý kcal mol21 (interaction energy of the hydrogen bonded dimer of water)1 will be considered as weak those with energies between ý25ý and ý210ý kcal mol21 will be defined as medium and those with energy values larger than ý210ý kcal mol21 will be defined as strong or very strong.Table 1 Non-conventional HBDs and HBAs—the object of the present review A–H···H–B Dihydrogen bonds (DHBs) A–H···B Hydrogen bonds (HBs) Protic Ref. 14 15 B = isonitriles10 B = carbanions11 B = carbenes12 B = p-systems10,13 2 2 Solvent and + electric field A–H···B?A···H–B + A–H···H–B?A···H–H···B effects Ref.17 Ref. 16 2 Hydric (inverese) + A–H···B?A–H···B Ref. 18 The development of new approaches to the treatment of electron density and its significance in the nature of chemical bonds have opened up new perspectives in the description of HBs making their definition more accurate and precise. Thus according to Koch and Popelier,4 and within the frame of the theory of atoms in molecules (AIM) proposed by Bader,19 the new additional criteria for a more correct definition of a HB will be c (i) Charge density (rc) and the Laplacian of the charge density (“2rc) at the bond critical points (BCPs see Appendix). The values obtained for these two parameters (small r values and positive “2rc values) should correspond to what is defined as ‘closed-shell’ interactions19 of the HB type (van der Waals complexes would have smaller rc values).(ii) Topology. By analyzing the bond paths between the H atom and the HBA it will be possible to determine the existence of an interaction such as a HB. (iii) Mutual penetration of the H and the interacting bond. In order to estimate the mutual penetration of the H atom and the HBA (r0 H and r0 A) upon HB formation the nonbonded radii of both parts have to be compared to the corresponding bonded radii (rH and rA BCP radii). The nonbonded radius is defineducleus to a 0.001 au charge density contour in the direction of the HB. Moreover if all the penetrations are positive they can be designated as HBs.(iv) Loss of charge and energetic destabilization of the H atom and total charge transferred. A necessary criterion for the formation of a HB is the loss of charge of the H atom involved. This loss (DN) is computed by subtracting the electronic population of the H in the free monomer from the corresponding H in the complex and should be negative. In addition this H atom should be destabilized in the complex and the destabilization is given by the difference in total atomic energy between the complex and the monomer (DE) which should be positive. Chemical Society Reviews 1998 volume 27 164 Finally the total charge transferred in the formation of the HB should always be negative implying the donation of electrons from one molecule to another.(v) Dipole moment enhancement. The electric charge rearrangement that accompanies the formation of a hydrogen bonded complex is an important characteristic of HBs. The difference between the dipole moment of the complex and the sum of those of the separate monomers provides some information of the electric rearrangement. These enhancements have contributions from the polarization of one monomer by the other and from the charge transferred in the formation of the HB. Thus the dipole moment of the complexes should be larger than the vector sum of the dipole moment of the monomers as a consequence of the HB formation. Taking into account all these conditions proposed to prove the existence of a HB let us survey the different nonconventional hydrogen bonded systems that are the subject of the present review.2 Isocyanides and carbon monoxide as HB acceptors In 1962 Ferstanding20 and Schleyer and Allerhand21 both described the case of an HB of the type C–H···C. The experiments showed that the carbon atom of isocyanide 2a R = Ph (also called isonitrile) acting as a HBA (HNC···HA) was able to form strong HBs with a variety of HBDs including carbon derivatives such as phenylacetylene. + O C H N C H H N C R N C H A 2d 2b 2a 2c We have studied several HB complexes 2b of hydrogen isocyanide (2a R = H) as a HBA with five well known HBDs (HF HOH HNH2 HCN) as well as with hydrogen isocyanide itself (HNC).10 Calculations were carried out at the MP2/ 6-311++G** and B3LYP/6-311++G** levels (see Appendix) and one example of the results obtained with the help of the AIM methodology19 is represented in Fig.1. The main Fig. 1 Relief map of the electron density (r)19 corresponding to the complex FH···:CNH conclusions were (i) concerning the geometry of the linear complex there is a lengthening of the H–N bond and a shortening of the N·C bond indicating for the isonitrile fragment of 2b a geometry intermediate between 2a and 2c and justifying our view of HBs as intermediates in protonation reactions; (ii) the interaction energies are significant (EI+BSSE = 25.4 kcal mol21 for hydrogen isocyanide to compare with 23.3 kcal mol21 for its isomer hydrogen cyanide) making these HBs of a weak to medium strength; (iii) the calculated chemical shifts (GIAO)22 are linearly correlated with the electron transfer in the complex; (iv) the X-ray structures found in the Cambridge Crystallographic Data Centre (CCDC),23 although scarce are well reproduced by the calculations.Concerning carbon monoxide 2d it has been established that the protonation in the gas phase occurs both on the carbon [HCO]+ and on the oxygen [COH]+.24 When acting as a HBA, however it uses the carbon and not the oxygen end24,25 unless the carbon is coordinated to a metal.26 Our calculations10 show the existence of weak HB (from 20.5 to 23.4 kcal mol21) between carbon monoxide and different HBDs. In contrast carbon monosulfide (CS) forms much stronger HBs (from 21.2 to 27.3 kcal mol21) which can be classified as weak to medium HBs.These differences can be explained based on the permanent dipole moment of both molecules. 3 Carbanions and zwitterions as HB acceptors Theoretical work has proposed that carbanions could act as HBAs.11 Thus Platts et al. have shown that carbanions stabilized as zwitterionic species 3a can form medium HBs with weak HBDs as acetylene (EI+BSSE = 27.7 kcal mol21) giving rise to complex 3b containing one of the less studied types of HBs—the C–H···C bond.11 The optimized geometry of 3b at the MP2/6-311++G** level shows a short C···H distance and the possible existence of a secondary interaction between the hydrogen atoms on nitrogen and the triple bond of acetylene. H H H C H H H H C N + 2.19 Å – H N H 84° H H 3a 3b C C H The strong basicity of the HBAs treated in the present section indicates that they would only be able to form HB with weak proton donors because more acidic donors would produce a spontaneous transfer of the hydrogen.4 Carbenes and silylenes as HB acceptors In this section we will discuss three related structures (i) carbenes 4a (ii) silylenes 4b and (iii) carbynes 4c. Si C C 4c 4b 4a We previously discussed the case of carbenes 4a and silylenes 4b and here we summarize only the main conclusions. 12 Although carbenes are in general highly reactive species with short lifetimes some examples are known of stable carbenes at room temperature.27 Silylenes have similar structures and were reviewed recently.28 Arduengo et al.29 also reported an X-ray structure of a stable ‘nucleophilic’ carbene that shows a new type of intermolecular HB 4d.It involves a carbonium and a carbene with an almost linear HB Mes Mes N Mes N N Mes N H H N N N Mes Mes Mes Mes 4d 3 + HCN HF H2O and HBAs include CH2 (singlet In our paper,12 a model (H2C:···H–CH2 +) of the structure reported by Arduengo was theoretically studied and this new type of HB was generalized in two ways (i) other HBDs than C+–H; and (ii) other HBAs than methylene. HBDs include CH methylene) SiH2 (singlet silylene) and CF2 (singlet difluorocarbene). The highest level of calculations used was the MP2 or MP4/6-311++G**. Quantitatively all the methods indicate the presence of HBs due to the short X:···H distances and interaction energies between 22 and 222 kcal mol21 depending on the complex studied.The values described by Arduengo [C:···H = 2.026(45) Å C:···H–C = 172.5° and H–C = 1.159(45) Å] for this new kind of HB are very similar to the ones obtained for the simplified model H2C:···CH3 + (C:···H = 1.988 C:···H–C = 180.0° and H–C = 1.102 Å). The model system shows a very strong HB (ca. 220 kcal mol21). For the strongest neutral complex (H2C:···HF Fig. 2) the interaction energy amounts to 210 kcal mol21. Fig. 2 Relief map of the electron density (r)19 corresponding to the complex FH···:CH2 The analysis of the charges gathered in Table 2 shows a flow of electrons from the atoms attached to the carbene or silylene (X) that become more positive to the more electronegative atom of the HBD (A).Table 2 Atomic charges (|e|) of the carbene and sylylene monomers and the HB complexes at the MP2/6-311++G** level using the AIM methodology HBAs B X 20.016 1.453 1.297 H2C H2Si F2C 0.008 20.726 20.648 HB complexesa DB DX 20.010 20.024 20.100 20.006 20.084 0.046 H2C:···HF H2C:···HCN H2C:···H2O H2C:···H3N H2C:···H3C+ H2Si:···HF H2Si:···HCN F2C:···HF 0.045 0.026 0.069b 0.012 0.084 0.008 0.006 0.012 0.007 0.014 0.024 0.000 F2C:···HCN a The atomic charges are relative to the monomers values. b Average of the two protons. In addition several relationships between the parameters that define these HBs (such as HB distance electron transfer electron density on the HB critical point and interaction energy) were found.Finally although the case of carbynes 4c was examined their multiplet nature complicates the calculations considerably.10 5 p Acceptors There has been much interest in hydrogen bonds where the HBA is a p-system especially acetylenes and benzenes.8 Thus Mingos et al.30 have characterized by X-ray crystallography the T-shaped intermolecular Cl3C–H···p (C·C) interactions present in chloroform solvates of gold(i) ethyne complexes. Recently Chemical Society Reviews 1998 volume 27 HBDs A H 0.712 0.204 0.567 0.348 0.275 20.712 0.817 21.133 21.043 0.176 HF HCN H2O H3N H3C+ Electron transfer DA DH 20.073 20.047 20.018 20.037 20.093 20.041 20.007 0.065 20.010 0.040 0.098 20.021 0.018 0.080 0.028 0.038 0.018 0.084 0.062 0.026 0.048 0.014 20.002 0.030 0.022 20.046 20.023 165 4H5 + is much more stable than the base Chandra et al.31 reported theoretical studies of diacetylene···HF and allene···HX (X = Cl F) complexes at the MP2 level of the theory and Chandra and Nguyen explored the possibilities of the same diacetylene as an HBA towards HCl and announced a similar study for the complex acetylene···HCl.31 The information about benzenes as p-HBAs is abundant microwave experiments have provided information about the spectrum and geometry of a series of HF or HCl complexes with benzene.32 Several crystal structures have been determined where these kinds of H···p interactions are observed intra-33 or intermolecularly.34,35 Tang et al. have reported a quantum chemical study (MP2/6-31G**//6-31G*) on a selection of p-type hydrogen-bonded systems.13 They used hydrogen fluoride as HBD and a set consisting of acetylene ethylene cyclopropane and benzene as HBAs. They were able to correctly describe a number of complexes previously identified by microwave spectroscopy. Moreover using the Bader topological approach, 19 they made a detailed description of the HBs. Due to a multitude of relevant experimental results appearing between the years 1991–1996 we decided to undertake a similar study10 also using HF as HBD but with two differences namely we used a 6-311++G** basis set and a large number of HBAs (the four of Tang plus cyclopropene cyclobutadiene and tetrahedrane).Tetrahedrane was selected because it is not only considered a three-dimensional aromatic system but also a strong base (the cation C C4H4).36 For all these complexes we have carried out the electron density analysis proposed by Koch and Popelier described previously4 representing in Fig. 3 the structure of two of these complexes Fig. 3(a) corresponds to the benzene···HF complex and Fig. 3(b) to the tetrahedrane···HF complex. In the case of the benzene complex [Fig. 3(a)] our minimum energy structure is Fig. 3 Geometries obtained at the MP2/6-311++G** level for the minimum energy H···p complexes formed between hydrogen fluoride and (a) benzene (b) tetrahedrane.Also shown are the distances between the H atom and (a) the p cloud (b) the C–C bond critical point. Chemical Society Reviews 1998 volume 27 166 different from that of Tang et al. (point symmetry C6v)13 and is more in agreement with experimental results.10,13 6 Dihydrogen bonds and electric field effects on dihydrogen bonds We14 and others15,37,38 have devoted some effort to the understanding of dihydrogen bonds (DHBs). These bonds mainly present in transition metal compounds are of the type ‘proton-hydride’ i.e. between an A–Hd+ and a B–Hd2. The main characteristics of the M–H···H–A systems (M = metal such as Ir or Re A = O N) are (i) close H···H contacts (1.75–1.90 Å) (ii) interaction energies in the range of weak conventional HBs (23 to 27 kcal mol21) (iii) large couplings (1JHHA = 2–4) between the AH (A = O) and the MH (M = Ir) protons and (iv) abnormally low minimum T1 values in the 1H NMR spectra.Experiments using complex 7 reveal that the N–H···F–Ir bond (25.2 kcal mol21) is a little stronger than the N–H···H–Ir DHB (25.0 kcal mol21).38 N H H H N N N H N H L L F H Ir F H Ir L L H H B H 7c 7b 7a These bonds were first generalized to the system B–H···H–N (7c).37 Theoretical calculations (PCI-80/B3LYP) sustained by structures found in the CSD show these bonds to be of medium strength (26.1 kcal mol21 for H3BNH3) and directional (H···H– N almost linear and B–H···H bent in the range 95–120°).The B–H···N bending is due to an attractive Coulombic interaction between the strongly negatively charged B and the protonic NHd+. We showed that DHBs are more general and that not only B–H···H–N (7c) but also several other systems present interaction energies EI+BSSE (in kcal mol21) charge densities at the hydrogen bond critical points rc (in e a023) and geometries which are consistent with the existence of a HB.14 The systems studied were BH42···HCN BH42···CH4 LiH···NH4 + LiH···HCN LiH···HCCH BeH2···NH4 + BeH2···HCN and CH4···NH4 +. An empirical model relating EI+BSSE to Mulliken populations predicts a EI+BSSE = 2106.5 kcal mol21 for the complex H3B2–H···H–NH3 +.The situation is not stable because when the structure is minimized it evolves to H3B + H2 + NH3. Contreras et al.38 have described experimental results (X-ray for the solid state and 1H NMR for the solution) consistent with the existence of ‘protic–hydric’ C–Hd+···d2H–B and protic–fluoride C–Hd+···d2F–B interactions (here C and B stand for carbon and boron). As a logical consequence of two of our lines of research (that on DHBs and that on the effect of electric fields16) and in order to investigate the hydrogen transfer within a crystal (by means of the crystal field) we studied the effect of an external field over three different proton equilibria with ab initio methods (HF/6-31G**).17 The equilibria chosen were eqn. (1)–(3). (1) (2) (3) H HBe–H···H–NH Li–H···H–F"Li+···H–H···F2 3N···H–H···BH3"H3N+–H···H–BH32 3 +"HBe+···H–H···NH3 An electric field in positive and negative directions was applied along the molecular axis increasing from 0.00557 to 0.03342 au in steps of 0.00557 au.In the absence of an external field structures in the left side of the equilibria were stationary points [with two negative frequencies in equilibria (1) and (2) and zero negative frequencies in equilibrium (3)]. Thus when applying an electric field (positive or negative) over the system on the right in equilibrium (1) this equilibrium is displaced to the left yielding a DHB system. Reciprocally when an increasing positive field is applied over the DHB system on the right of the equilibrium (2) we obtained the system on the left.In the complexes with coordinated hydrogen molecules of equilibria (2) and (3) the H–H bonds were broken by the field effect resulting in two molecules in each case that moved to the infinite. Further by applying an increasing electric field it was possible to keep connected an unstable DHB complex [right side of equilibrium (2)] or an unstable hydrogen coordinated complex [right side of equilibrium (3)]. In conclusion we found that by applying an external electric field the transfer of H atoms between two heavy atoms is possible as it occurs within the crystal due to internal forces. Recently a situation corresponding to the right side of eqn. (1) has been described namely CsF(H2) and KF(H2).39 7 Inverse HB complexes In all the unusual HBs mentioned above the H atom plays the role of electron acceptor with the exception of the DHBs where one of the H atoms accepts the electrons while the other provides them.Following this sequence we studied a new class of unconventional HBs where the H atom provides electrons and another non-hydrogen atom accepts them.18 e H A B e H H A B e A DH DX(1,2) a H Li–H···Li–H H–Be–H···Li–H 0.022 0.011 0.013 0.010 0.013 0.160 20.015 0.009 20.010 20.010 20.036c B In order to obtain these ‘inverse’ HBs we should consider a particular set of molecules formed by ‘donors’ and ‘acceptors’ of electrons. On one hand LiH BeH2 and BH42 with very electron deficient heavy atoms will be electron donors (‘e-donors’).On the other electron acceptors (‘e-acceptors’) will be the Li or Be hydrides because these alkaline atoms would accept the electrons easily and without any steric restriction. Besides other Li and Be derivatives with methyl or fluoride groups have been included in the ‘e-acceptors’. The complexes studied were (i) Those formed by ‘linear approximation’ between a Li Be or B hydride (‘e-donors’) and a Li derivative (‘e-acceptors’). (ii) Those formed by ‘multiple approximation’ between Li and Be hydrides and fluorides where the metal hydrogen and/or fluorine atoms form a larger number of linkages than their corresponding valence. This group has been studied only for comparative purposes because the interactions involved cannot be properly considered as HBs.Many of the structures discussed in ref. 18 both monomers and complexes have not been observed but compounds of Li Be and B that are experimentally known span a wide range of structural types.18 Some of these complexes were previously calculated. For example in the case of the linear (LiH)2 dimer some authors termed the interaction ‘lithium bond’ and other authors considered these aggregates as bounded binary com- Table 3 Atomic charges (ýeý) s of the inverse HB linear complexes at the MP2/6-311++G** level using the AIM methodology 3 H–Be–H···Li–CH H–B·(H)3···Li–H a Atoms attached to the H atom in order from the nearest to the farthest. b Atoms attached to the Li atom in order from the nearest to the farthest.c Average of the three protons. plexes. None however suggest these interactions to be inverse HBs.18 Given the basically accepted definition of a HB—a bond distance d(H···A) shorter than the sum of the atomic van der Waals radii of H and A; a bond angle a(B–H···A) almost linear; a certain transfer of charge among these three atoms and an energy around 22 to 222 kcal mol21—our study suggested that linear complexes fulfil all the conditions and therefore should be considered as inverse HBs. The computation of the atomic charges of the linear complexes showed that there is an electron transfer from fragments defined as ‘e-donors’ to the ‘e-acceptors’ in all the complexes studied as can be seen in Table 3. Therefore and contrary to classical HBs the charge and the H atom flow in the same direction from the acid to the basic fragment.This makes these inverse HBs unique. Within the frame of the AIM theory,19 the electronic distribution in these inverse HBs shows a hydrogen atom bound to both the ‘e-donor’ and the ‘e-acceptor’ by closed-shell interactions both of which are closed-shell interacting fragments. The bond critical points obtained in these interactions reflect all the characteristics associated with HBs low rc values and “2rc > 0. The energy results showed that the inclusion of correlation effects diffuse functions and ZPE and BSSE corrections are significant in the description of these particular HB complexes. Linear complexes of BeH2 showed interaction energies within 25 to 210 kcal mol21 whereas the LiH dimer and the BH42 complex exhibited interaction energies around 225 to 250 kcal mol21.8 Conclusions (i) Although much of our contribution is related to calculations of small systems that is to idealized ‘gas-phase’ simple situations according to ref. 5 ‘The study of HB interactions in the gas-phase is relevant because it provides direct information on the structure and stability of these complexes in the absence of the perturbations induced by the solvent’. In the solid-state (mostly X-ray structures) care should be taken not to define HBs only in geometrical terms following Dunitz’s advice it is important not to confuse attractive–repulsive interaction forces with stabilizing–destabilizing interaction energies.40 Moreover a short distance (H···B or H···H) alone is not proof of the existence of a HB or DHB.(ii) Another point worthy of consideration is that HBs can be viewed as intermediates in protonation processes.16 For instance DHBs may lie on the pathway whenever a hydride undergoes protonation14,37 eqn. (4). M–H2 + H+–A?M–H···H–A?M+ H2 + A (4) The problem of the relation between HBs and protonation is very general being found in the three states of matter in the solid state where a continuum of situations between A–H···B and A2···H–B+ is observed in X-ray crystallography and in solid state NMR by slightly modifying the nature of A and B;16 in the gas-phase where linear relationships between thermodynamic proton affinities and HBs have been reported;6 and in solution where the study of the complex relations between both properties is the most developed.5 Electron transfer (1,2) b DY DLi 0.031 0.014 0.013 0.037 20.010 20.003 0.020 20.008c 20.022 20.021 20.011 20.009 20.015 167 Chemical Society Reviews 1998 volume 27 (iii) In 1970 Abboud and Bellon and Sherry and Purcell simultaneously proposed models of the form EHB = a b a and b being parameters related to HB acidity and basicity.5 We have reported in Table 4 some values of -EI+BSSE from our publications.10,12 Table 4 Matrix of energies of different complexes (2EI+BSSE at the B3LYP/6-311++G** level) in kcal mol21 ·CH CS CO CNH HBA 3 :CH2 HBD 3.061 1.205 1.367 0.488 0.485 12.070 5.458 5.921 2.055a 2.681 7.253 3.450 3.254 1.236 1.551 3.356 1.253 1.324 0.454 0.568 5.763 3.891 2.849 1.148 1.813 FH NCH H2O NH3 HC·CH a Value estimated from the model E = a b (see Table 5).We have deconstructed these and other -EI+BSSE values into a product of two terms one characterizing the HBD (a) and the other the HBA (b) (assuming that the value of b for the methyl radical is 1). The results are reported in Table 5 and constitute a scale of a and b values for the gas-phase the largest values corresponding to the strongest HBDs (FH) and HBAs (:CH2). Table 5 a and b values corresponding to the HB energies of Table 4 HBA (b) HBD (a) 2.773 1.285 1.256 0.459 0.579 3 FH NCH H2O NH3 HC·CH 2.564 1.038 2.649 4.477 1 1.432 1.475 1.251 1.248 1.598 1.358 1.145 CNH CO CS :CH2 ·CH HC·CH CH2NCH2 Cyclopropene Cyclopropane Cyclobutadiene Tetrahedrane Benzene (iv) In this review we have tried to present an integrated view of hydrogen bonds.Using the ‘protic’/‘hydride’ nomenclature, 37,38 HBs can be classified in three groups protic HBs hydric HBs and protic-hydric DHBs. In the case of protic HBs non-conventional HBDs correspond for instance to the case when A is a carbon atom (C–H···B B = O N),7–9 while several non-conventional HBAs have been reviewed here (isonitriles carbanions carbenes and p-systems). We have summarized the results concerning protic–hydric DHBs and those of the new class of hydric inverse HBs.A simple picture seen below could represent the situation of hydrogen bonded complexes today. Obviously if the area were proportional to the importance (or the number of references) then protic HBs would cover more than 99.9% of the pie surface (Fig. 4); we expect that this Fig. 4 situation will become more balanced in the future. This review has tried to demonstrate that non-conventional hydrogen bonds Chemical Society Reviews 1998 volume 27 168 although in some cases weak are still one of the most important of the non-covalent interactions. 9 Appendix In this section the technical terms used to describe the calculations performed over the systems reviewed here will be defined to aid understanding.9.1 Molecular orbital ab initio calculations This approximate treatment of electron distribution and motion assigns individual electrons to one-electron functions. These contain a product of spatial functions termed molecular orbitals y1(x,y,z) y2(x,y,z) . . . (MOs). In the simplest version of the theory a single assignment of electrons to orbitals is made. These orbitals form a many-electron wavefunction Y which is the simplest MO approximation to the solution of the Schr�odinger equation. In practical calculations the molecular orbitals y1 y2 . . . are considered as a linear combination of a set of N known one-electron functions f1(x,y,z) f2(x,y,z) . . . eqn. (5). N i m fm i (5) m 1 = Â c Y = The f1 f2 .. . functions are known as one-electron basis functions and they constitute the basis set. When these basis functions are the atomic orbitals for the atoms of the molecule eqn. (5) is described as the linear combination of atomic orbitals (LCAO) approximation. 9.2 Hartree–Fock (HF) approximation Imposes two constrains in the resolution of the Schr�odinger equation and to obtain the energy (i) the use of a limited basis set in the orbital expansion and (ii) the use of a single assignment of electrons to orbitals. 9.3 Møller–Plesset (MP) perturbation energy This is an alternative approach to the correlation problem. Within a given basis set it tries to solve the full Hamiltonian matrix (within the Schr�odinger equation).The approach is to treat the matrix as the sum of two parts the second being a perturbation of the first. 9.4 MP2 MP4 By carrying out MP2 MP4 the energy can be expressed as series and practical correlation methods may be formulated by truncation of the series to various orders. We refer to the methods by the highest-order energy term allowed thus MP2 means that the truncation has been made after the second-order and MP4 after the fourth-order. 9.5 Differential functional theory (DFT) In DFT the exact exchange of the Hartree–Fock method for a single determinant is replaced by a more general expression the exchange-correlation function which can include terms accounting for both exchange energy and the electron correlation expressed as a function of the density matrix which is omitted from the HF theory.9.6 B3LYP B3LYP is a hybrid method which includes a mixture of the HF exchange with DFT exchange-correlation. This functional described as Becke3 (B3) is the three-parameter exchange functional containing Slater exchange functional HF and Becke’s 1988 gradient correction and the LYP (Lee–Young– Parr) correlation functional. 9.7 6-31G* and 6-31G** These are commonly used split-valence plus polarization basis sets which contain inner-shell functions each written as a linear combination of six gaussians and two valence shells represented by three and one gaussian primitives respectively (represented by 6-31G). In addition a set of six d-type gaussian primitives has been added to each heavy atom (represented by *) and a single set of gaussian p-type functions to each hydrogen atom (represented by **).9.8 6-311++G** This is a split-valence basis set plus polarization and diffuse functions. It comprises an inner shell of six s-type gaussians and an outer (valence) region which has been split into three parts represented by three one and one primitives respectively (represented by 6-311G). The basis is supplemented by a single set of five d-type gaussian functions for first-row atoms and a single set of uncontracted p-type gaussians for hydrogen (represented by **). In addition it incorporates two sets of diffuse gaussian s- and p-type functions (represented by ++). 9.9 Mulliken population analysis This is an analysis of the electron population of how a molecule distributes electrons according to the atomic orbital occupancy and the overlap population between two atoms is arbitrarily divided evenly between both of them without taking into account possible differences in coefficients atom types electronegativities and other aspects.9.10 EI+BSSE With this expression is defined the interaction energy of a HB complex corrected by the basis set superposition error (BSSE). BSSE refers to vacant orbitals on one atomic centre being used to make up for a basis set deficiency on a neighboring atom. The interaction energies of all the complexes are calculated as the difference between the total energy of the complex and the total energy of the isolated monomers (EI = EAB 2 {EA + EB}).Because the computed interaction energies will be affected by the basis set superposition error (BSSE) the latter has been estimated using the full counterpoise method and eqn. (6) where BSSE (A–B) = E(A)A2E(A)AB + E(B)B2E(B)AB E(A)AB represents the energy of the monomer A calculated using its geometry within the dimer and the complete set of basis functions used to describe the dimer; and E(A)A is the energy of the same molecule but using only the basis functions centred on it. 9.11 Gauge independent atomic orbital (GIAO) It represents the eigenfunctions of the one-electron system that have been perturbed by an external magnetic field. 9.12 AIM methodology According to Bader et al.19 the theory of atoms in molecules (AIM) offers a self-consistent way of partitioning any system into its atomic fragments considering the gradient vector field of its electron density r.By means of a topological analysis features such as critical points and paths of maximum electron density (atomic interaction lines) can be studied since AIM provides a ‘molecular graph’ which is a representation of the bonding interactions. 9.13 Electron density function (r) EDF is a three-dimensional function defined such that r(r) dr is the probability of finding an electron in a small volume element dr at some point in the space r. 9.14 rc This is the electron density found at the bond critical point (BCP which represents the point between two atoms with a minimum in the path of maximum electron density).c 9.15 “2r This is the Laplacian of the electron density found at the bond critical point. 10 Acknowledgements This review owes much to discussions with Professors H.-H. Limbach (Berlin) and M. Y�a�nez (Madrid) and Drs B. Chaudret (Toulouse) O. M�o and M. Alcam�ý (Madrid) within the framework of the EU network ‘Localization and Transfer of Hydrogen’ (No. CHRX CT 940582). We thank Professor Jack D. Dunitz (Z�urich) for allowing us to quote his opinion (ref. 40) before publication. References 1 F. Hibbert and J. Emsley Adv. Phys. Org. Chem. 1990 26 255. 2 J. C. MacDonald and G. M. Whitesides Chem. Rev. 1994 94 2383. 3 J. Bernstein M. C. Etter and L. Leiserowitz The Role of Hydrogen Bonding in Molecular Assemblies in Structure Correlation ed.H.-B. Dunitz VCH Weinheim 1994 p. 431. 4 U. Koch and P. L. A. Popelier J. Phys. Chem. 1995 99 9747. 5 J. L. M. Abboud R. Notario and V. Botella Quantitative Treatments of Solute/Solvent Interactions Theoretical and Computational Chemistry 1994 1 135 and references therein. 6 J.-F. Gal and P.-C. Maria Prog. Phys. Org. Chem. 1990 17 159 and references therein. 7 G. R. Desiraju Acc. Chem. Res. 1996 29 441. 8 T. Steiner E. B. Starikov A. M. Amado and J. J. C. Teixeira-Dias J. Chem. Soc. Perkin Trans. 2 1995 1321. 9 P. Seiler L. Isaacs and F. Diederich Helv. Chim. Acta 1996 79 1047. 10 I. Alkorta I. Rozas and J. Elguero Theor. Chem. Acc. 1998 in press and references therein accepted with modifications; I.Rozas I. Alkorta and J. Elguero J. Phys. Chem. A 1997 101 9457 and references therein. 11 J. A. Platts S. T. Howard and K. Wozniak Chem. Commun. 1996 63; J. A. Platts and S. T. Howard J. Chem. Soc. Perkin Trans. 2 1997 2241. 12 I. Alkorta and J. Elguero J. Phys. Chem. 1996 100 19367. 13 T.-H. Tang W.-J. Hu D.-Y. Yan and Y.-P. Cui Theochem. 1990 207 319. 14 I. Alkorta J. Elguero and C. Foces-Foces Chem. Commun. 1996 1633. 15 Q. Liu and R. Hoffmann J. Am. Chem. Soc. 1995 117 10108. 16 M. Ramos I. Alkorta J. Elguero N. S. Golubev G. S. Denisov H. Benedict and H.-H. Limbach J. Phys. Chem. A 1997 101 9791. 17 I. Rozas I. Alkorta and J. Elguero Chem. Phys. Lett. 1997 275 423. 18 I. Rozas I. Alkorta and J. Elguero J. Phys. Chem. A 1997 101 4236.19 R. F. W. Bader Atoms in Molecules. A Quantum Theory Oxford University New York 1990. 20 L. L. Ferstanding J. Am. Chem. Soc. 1962 84 1323 and 3553. 21 P. von R. Schleyer and A. Allerhand J. Am. Chem. Soc. 1962 84 1322 and 1963 85 866. 22 GIAO Gauge Including Atomic Orbitals D. B. Chesnut Ab Initio Calculations of NMR Chemical Shielding in Ann. Rep. NMR Spectrosc. 1994 29 71. 23 CSD F. H. Allen J. E. Davies J. J. Galloy O. Johnson O. Kennard C. F. Macrae E. M. Mitchell G. F. Mitchell J. M. Smith and D. G. Watson J. Chem. Inf. Comput. Sci. 1991 31 187. 24 P. C. Burgers J. L. Holmes and A. A. Mommers J. Am. Chem. Soc. 1985 107 1099. 25 A. C. Legon and D. J. Millen Chem. Rev. 1986 86 635 and references therein. 26 D. Braga and F. Grepioni Acc. Chem. Res. 1997 30 81. 27 M. Regitz Angew. Chem. Int. Ed. Engl. 1996 35 725. 28 M. Driess and H. Gr�utzmacher Angew. Chem. Int. Ed. Engl. 1996 35 828. 29 A. J. Arduengo J. R. Goerlich and W. J. Marshall J. Am. Chem. Soc. 1995 117 11027. 30 T. E. M�uller M. P. Mingos and D. J. Williams J. Chem. Soc. Chem. Commun. 1994 1787. 31 R. Sumathi and A. K. Chandra Chem. Phys. Lett. 1997 271 287; A. K. Chandra and M. T. Nguyen J. Chem. Res. (S) 1997 216 and references therein. 169 Chemical Society Reviews 1998 volume 27 32 J. A. Shea R. E. Bumgarner and G. Henderson J. Chem. Phys. 1984 80 4605. 33 J. Wilamowski D. Osman J. J. Sepiol and N. Rodier Aust. J. Chem. 1996 49 951. 34 M. A. Viswamitra R. Radhakrishnan J. Bandekar and G. R. Desiraju J. Am. Chem. Soc. 1993 115 4868. 35 T. Steiner J. Chem. Soc. Chem. Commun. 1995 95. 36 R. Notario and J. Elguero J. Chem. Soc. Chem. Commun. 1995 1543. 37 R. H. Crabtree P. E. M. Siegbahn O. Eisenstein A. L. Rheingold and T. F. Koetzle Acc. Chem. Res. 1996 29 348. Chemical Society Reviews 1998 volume 27 170 38 I. I. Padilla-Mart�ýnez M. J. Rosalez-Hoz H. Tlahuext C. Camacho- Camacho A. Ariza-Castolo and R. Contreras Chem. Ber. 1996 129 441. 39 R. L. Sweany and J. S. Ogden Inorg. Chem. 1997 36 2523. 40 J. D. Dunitz Plenary Lecture 13th International Conference on the Chemistry of the Organic Solid State Stony Brook New York USA July 13–18 1997. Received 21st May 1997 Accepted 3rd Novembe

ISSN:0306-0012    

DOI:10.1039/a827163z

出版商:RSC    

年代:1998

数据来源: RSC