Early Main Group Metal Catalysts for Imine Hydrosilylation Holger Elsen,Christian Fischer, Christianknüpfer,Ana Escalona, and Sjoerd Harder*[A]
Total Page:16
File Type:pdf, Size:1020Kb
DOI:10.1002/chem.201904148 Full Paper & Catalysis |Hot Paper| Early Main Group Metal Catalysts for Imine Hydrosilylation Holger Elsen,Christian Fischer, ChristianKnüpfer,Ana Escalona, and Sjoerd Harder*[a] Abstract: The efficient catalytic reduction of imines with tivitiesincrease with metal size. For most substrates the ac- phenylsilaneisachieved by using the potassium, calcium tivity increases along the row K< Ca<Sr< Ba. Fastest con- and strontium based catalysts [(DMAT)K(THF)] , version was found for imines with alkyl substituents at N 1 (DMAT)2Ca·(THF)2 and (DMAT)2Sr·(THF)2 (DMAT= 2-dimethyl- and aryl rings at C, for example, PhC(H)=NtBu, while amino-a-trimethylsilylbenzyl). Eight different aldimines and tBuC(H)=NtBu or PhC(H)=NPh react much slower. Reason- the ketimine Ph2C=NPh could be successfully reduced by able functionalgroup tolerance is observed. The proposed PhSiH3 at temperatures between 25–608Cwith catalystload- metal hydride mechanism is supported by stoichiometric re- ings down to 2.5 mol%. Also, simple amides like KN(SiMe3)2 actions using acatalystmodel system, isolation of intermedi- or Ae[N(SiMe3)2]2 (Ae =Ca, Sr,Ba) catalyze this reaction. Ac- ates and DFTcalculations. Introduction especially the metals potassium and calcium hold the advan- tage that they are both non-toxic as well as inexpensive, they Due to the high demandofamines in the field of pharmaceuti- make for excellent alternatives to noble-metal catalysts. Asgari cals and agrochemicals, the catalytic conversion of imines to et al. reported the hydrosilylation of activatedalkenes using amines has become an important field of research.Although a simple s-block metal bases such as KOH or alcoholates.[9] This, large part of this research focusesondirect imine hydrogena- however,has the drawback that the high degree of regiocon- tion, the reduction of the C=Nbond with an easy to handle trol observed by Harder is lost. In 2008 we extended our work and safe silane has become an important alternative.[1] While with group 2metal catalysts for ketone hydrosilylation.[10] Espe- late transition metal catalysts are well-established for imine hy- cially for ketonesubstrates like RCH2C(O)R’,which easily form drosilylation,[2,3] the focus of current researchisshifted towards enolatesbydeprotonation, amismatch was found between cheaper more environmentally and benignmetals.[4,5] In 2006, products obtained from stoichiometricreactions with acalci- Yunetal. published the first catalytic hydrosilylation of imines um hydride complexand those obtained in acatalytic hydrosi- using azinc based catalyst.[6,7] There is, however,little known lylation regime. On this basis we proposedamechanism that in literature about the catalytic hydrosilylation of C=Nbonds circumvents formation of ametal hydride intermediate and in- using s-block metals. stead it was suggestedthat the hydride is transferred directly In 2006, we introduced alkenehydrosilylation using potassi- from asilicate intermediate (Scheme 1b). Later reports on um-, calcium- and strontium-based catalysts (Scheme 1a)and ketonehydrosilylation with simple potassium alkoxide catalysts found that the regioselectivity in some cases can be tuned support this hypothesis and propose asimilar mechanism.[11] either by metal or solventchoice.[8] Formation of the anti-Mar- Although early main-group-metal catalysis has made sub- kovnikoff product was explained by the hydride cycle whereas stantialprogress,[12] it is surprising that hitherto imine hydrosi- for the Markovnikoff product asilanidecycle is operative. Since lylation has not been reported.The closest to imine hydrosilyl- ation is the recently described catalytic dearomatization of the C=Nbond in pyridine by 1,2-selective hydrosilylation using a [a] H. Elsen, C. Fischer, C. Knüpfer,Dr. A. Escalona, Prof. Dr.S.Harder [13] Inorganic and Organometallic Chemistry calcium catalyst. In contrast, magnesium catalysts showed Friedrich-Alexander-UniversitätErlangen-Nürnberg no activity in pyridine hydrosilylation but performed well in its Egerlandstraße 1, 91058 Erlangen (Germany) hydroboration.[14,15] In this work we present the first investiga- E-mail:[email protected] tions on imine hydrosilylationusing early main group metal Homepage:www.harder-research.com catalysts. We optimize the reaction conditions, discussthe sub- Supporting Information (crystal structure data including ORTEP representa- tion, selected NMR spectra, details for investigations on reaction intermedi- strate scope and also provide experimental as well as calcula- ates, details and xyz coordinate files for the DFT calculations) and the tional evidencefor apotentialmechanism. ORCID identification number(s) for the author(s) of this article can be found under:https://doi.org/10.1002/chem.201904148. 2019 The Authors. Published by Wiley-VCH Verlag GmbH&Co. KGaA. Results and Discussion This is an open access article under the terms of the Creative Commons At- tributionLicense, whichpermits use, distributionand reproduction in any The hydrosilylationofimines has been investigated using the medium, provided the original work is properly cited. following complexesascatalysts (Figure 1): [(DMAT)K(THF)] 1 Chem. Eur.J.2019, 25,16141 –16147 16141 2019 The Authors. Published by Wiley-VCH Verlag GmbH &Co. KGaA, Weinheim Full Paper Scheme1.(a) Proposed mechanisms for the catalytic alkene hydrosilylation with K, Ca or Sr catalysts. Metal or solvent choices determine the mechanism and regioselectivity.(b) Proposed silicate cyclefor catalytic ketone hydrosilylation. (DMAT)2Ca·(THF)2 (2)convertedimine I significantly faster,both at room temperature and 608C(entries 5–6). The more polar and reactive Sr catalyst 3 was, as expected, clearly the fastest catalyst: at room temperature imine I was fully hydrosilylated by PhSiH3 within 5minutes using only acatalyst loading of 2.5 mol%(entry 8). Furtherloweringofthe catalystloading, however,led only to traces of product, even at 60 8C(entry 9). It is noteworthy that the simple amide complexes Figure 1. Catalysts for the hydrosilylation of imines. M[N(SiMe3)2]n (M= K(n= 1), Ca, Sr,Ba(n=2)) were also found to be catalytically active (entries 13–16) with activities increas- ing along the row K< Ca<Sr<Ba. Conversions with Ca[N- (1), (DMAT)2Ca·(THF)2 (2)and (DMAT)2Sr·(THF)2 (3). These com- (SiMe3)2]2 or Sr[N(SiMe3)2]2 are clearly slower than those with plexes, which are already well established catalysts for ketone the corresponding DMATcomplexes 2 and 3 but that of the [8,10] or alkene hydrosilylation, contain the 2-dimethylamino-a- Ba[N(SiMe3)2]2 catalyst is on par with that of Sr catalyst 3. trimethylsilylbenzylanion (DMAT) in which the negative charge Switching from benzene to the more polar solventTHF led is stabilized through negative hyperconjugation by the Me3Si to adrastic increase of the catalytic performance of the Kcata- substituent as well as by delocalization in the aromatic ring. lyst (entry 4). This may be explained by its much bettersolubil- The ortho-Me2N-substituent provides stabilization by metal co- ity in THF.Incontrasttothis, both the Ca and Sr catalysts need ordination.The Ca and Sr complexes are highly stable molecu- in THF significantly longerreaction times for imine hydrosilyl- lar entities that do not deprotonateTHF and are very well solu- ation (entries 7and 10). Apparently,the Lewis acidity of the ble in aromatic solvents like benzene. However,the Kcomplex metal is muchmore important in group 2metal catalysis than forms acoordination polymer in the solid state and shows for the potassium-catalyzed reaction. Thus, THF competes with only limited solubility in aromatic solvents. the imine for precoordination and substrate activation. All catalytic reactions were carried out in C6D6 and weredi- The influence of the silane on the conversion rates was also rectly monitored by 1HNMR (Table 1). Substrate conversion has investigated. Using Ca catalyst 2 (5 mol%) the benchmark been determined by integration of characteristic reactant and imine I was at room temperature reduced by PhSiH3 within product signals. Thehydrosilylation of avariety of imines (I–IX) 15 minutes(entry 5). Withthe secondary silane Ph2SiH2 after with PhSiH3 catalyzed by 1–3 has been investigated. In the ab- 24 hours only 9% conversion wasobserved. While the tertiary sence of acatalyst, the imine substrates did not react with silane Ph3SiH gave after 24 hours only traces of the hydrosilyl- phenylsilane. In most cases, the catalytic imine hydrosilylation ation product, silane (EtO)3SiH did not yield any conversion at proceeded cleanly and no side products were observed in all. For this reason,further investigations on the scope of 1 HNMR or GC-MSanalysis. imines have been performed with PhSiH3. In afirst series of experiments the benchmark imine sub- Functional group tolerance has been investigated for the III IV strate E-PhC(H)=NtBu (I)was converted with PhSiH3 in order to imines para-X-PhC(H)=NtBu in which X= MeO (II), Cl or Me . optimize the reaction conditions. The Kcatalyst 1 gave at Introduction of a para-substituent in the phenylring generally room temperature hardly any conversion (entry 1). Since this increases reaction times. The methoxysubstituent (imine II) may find its origin in the poor solubility of 1 in C6D6,asmall only exhibited asignificant effect for the Ca catalystincreasing amount of THF (ca. 2.5equiv) was added but this barely gave the necessary reaction time to 80 minutes (entry 18). Achloro any improvement. An increaseofthe catalyst loading from 2.5 substituent in