DOI:10.1002/chem.201602679 Communication

& Economy Self-SufficientFormaldehyde-to-Methanol Conversion by Organometallic FormaldehydeDismutase Mimic Dominic van der Waals+,[a] Leo E. Heim+,[a] Simona Vallazza,[a] ChristianGedig,[a] Jan Deska,[b] and MartinH.G.Prechtl*[a]

In nature, plays amuch more pronounced Abstract: The catalytic networks of methylotrophic organ- role within the familyofC1 molecules. Basedonanevolution- isms, featuring redoxenzymes for the activation of one- ary conserved detoxification mechanism, various methanol-tol- carbon moieties, can serve as great inspirationinthe de- erant or even methanol-feeding microorganisms have devel- velopment of novel homogeneously catalyzed pathways oped abiocatalytic machinery to deal with,and benefit from for the interconversion of C1 moleculesatambient condi- formalin. In addition to the capability to include formaldehyde tions. An imidazolium-tagged arene–ruthenium complex into the biosyntheticcarbon fixationbythe ribulose mono- was identified as an effective functional mimicofthe bac- phosphate pathway,[20,21] methylotrophs exploit formalin, rather terial formaldehyde ,which providesanew and than methanol, as asource of reduction equivalents. Therein, highly selectiveroute for the conversion of formaldehyde the preactivation of CH2Obythe formation of hemithioacetal to methanol in absence of any externalreducing agents. conjugates with either cofactors (mycothiol or glutathione)[22,23] Moreover,secondary amines are reductively methylated or protein-bound mercaptanes[24] allows for atransferhydroge- by the organometallic dismutase mimic in aredox self-suf- nationtoNAD+,inwhich NADH is liberated to serve as biolog- ficient mannerwith formaldehyde acting both as carbon ical reductant(Scheme1,top left). Inspiredbythis mode of source and reducing agent. action,werecently reported on abiomimetic ruthenium-based

H2 release system using methanediol as simple tetrahedral formaldehyde conjugate analogue and hydrogenasabiotic Methanoland formaldehyde are key platform chemicals that NADH equivalent (Scheme 1, bottom left).[25] [1–3] are industriallyformed from syngasonamegaton scale. Agreat number of C1-feeding bacterial strains,for example Currently,these reactionsare carriedout at high temperatures Pseudomonas putida, Staphylococcus aureus,orMycobacterium and pressures over variousdifferent heterogeneous cata- lysts.[4,5] Milder reaction conditions for the conversion of one- carbon entities have been achieved using well-definedmolecu- lar metal catalysts. Therein, the most successful examplescom- monly focussed on highly oxidized starting materials, such as carbon dioxide and formic acid. In addition to well-developed [6–11] CO2 to formatereduction protocols, both multi-metallic ap- proaches[12] and single-site catalystsystems[13–17] have emerged en route to the homogeneously catalyzed methanol synthesis from CO2 in the past five years.Moreover,the methanol pro- ductionwas attempted by catalytic disproportionation of formic acid. Fighting against the favorable formate decomposi- tion,[18] in 2014 aruthenium–triphos complex was reported to generate MeOH in up to 50%yield along with at least two [19] equivalents of CO2.

[a] Dr.D.van der Waals,+ L. E. Heim,+ S. Vallazza,C.Gedig, Dr.M.H.G.Prechtl Department für Chemie, UniversitätzuKçln Greinstrasse6,50939Cologne (Germany) E-mail:[email protected] Homepage:http://catalysis.uni-koeln.de/ [b] Prof. Dr.J.Deska Scheme1.a) The primary routes of the biological formaldehyde metabolism Department of Chemistry,Aalto-yliopisto proceed by dehydrogenation and disproportionation by two structurally Kemistintie 1, 02150 Espoo(Finland) linkednicotinamide-depending .b)Bioinspired ruthenium- [+] These authors contributed equally to this work. catalysed H2 release and methanol synthesis from hydrated formaldehyde. Supportinginformation for this article is available on the WWW under *Formate serves as secondreduction equivalent for the conversion of form-

http://dx.doi.org/10.1002/chem.201602679. aldehydetomethanol under liberationofCO2.

Chem. Eur.J.2016, 22,11568 –11573 11568 2016 Wiley-VCH Verlag GmbH &Co. KGaA, Weinheim Communication gastri,supplement their formaldehyde metabolism by asecond pathway using dismutases that are independent of externalsacrificialredox partners(Scheme1,top right).[26–28] On one side,theseformalin disproportionating exhibit aconsiderable structuraland functional resemblance of gluta- thione-independent zinc-containing dehydrogenases,withase- quencesimilarity greater 70 %(Supplementary Figure 1inthe Supporting Information).[29] As aresult of afirmly immobilized nicotinamide-dinucleotide inside the ,[30,31] however, initial dehydrogenation of the conjugated formalin is accompanied by areduction of asecond formaldehyde moiety.Hence, the deviated catalytic behavior provides both methanol and formate from two molecules of formaldehyde and water.Considering the analogy of NADH as biological hy- drogen carrier and our lately described bioinspired process fea- Scheme2.Proposed routes for the catalytic formaldehyde disproportiona- [25] turing an acceptorless H2 liberation, we envisioned that tion based on H2 coupled dehydrogenation/hydrogenation pathways or the modification of the organometallic species and/or the reaction dismutase-like by intermediary hydride species. [a] H2/HCO2Hcan environment of our original protocol will lead to anovel ho- potentially act as reducing agents for anotherequivalentofformaldehyde. mogeneously catalyzed formaldehyde-to-methanol converting system,(Scheme 1, bottom right) yet unprecedented in the Table 4) and, thus, cannot be considered asaproductive path- [32,33] context of abiotic C1-valorization pathways. way.Under the assumption thatfree H2 is involved as redox Recently,wereported on the possibility to incorporate mediator, two hydrogen-coupled routes are feasible. As metha- aruthenium-based formaldehyde dehydrogenase mimic into nediol has been shown to readily dehydrogenate to yield CO2 an artificial metabolism,which nicely cooperates with metha- and H2,itwas considered thatinaclosed system the direct re- nol-activating enzymes to provide aroom-temperature path- ductionofCO2 to methanol might be occurring (Scheme 2, way for the MeOH to H2 conversion and showcases the poten- left). However,neither precharging of the reactionvessel with [34] tial of chemoenzymatics in the small molecule activation. carbon dioxide nor the removal of superfluous CO2 by Ca(OH)2 However,inour crimp-top setup for the in situ gas phase anal- resulted in measurable effects on the methanol yields (Supple- ysis, the apparent turnover numbers of the H2 liberation in mentaryTable 5). This suggeststhat the direct CO2 reduction aqueous phosphate buffer laggedbehind the uncoupled might not be the primary mechanism for the methanol forma- system,whichcan in parts be attributed to infavourable tion. Additionally,the dismutation of 13C-labelled paraformalde- 12 metal–protein interactions, or might be relatedtothe reversi- hyde was conducted under elevated pressure of CO2 to get bility of the process by the formalin reductionunder elevated furtherinsightinto the potentialrole of carbon dioxide.While

H2 pressure. This finding served as astarting point forthe redi- the yields were not affected by the level of CO2 in the pressur- rection of the catalytic profile towards aformaldehyde dismu- ized atmosphere (Supplementary Table 6), the 1HNMR spectro- 13 tase mimicthat is disclosed in this communication. scopical analysis of the methanol obtained from the CH2O 12 To our delight, already slight modifications of our parent disproportionation under 15 bar CO2 provided no significant 12 [Ru(p-cymene)Cl2]2 catalysed H2 releaseprotocol, namely evidenceofthe CH3OH formation (Supplementary Figure 2), aclosed-vessel system and increased reaction temperatures, suggesting that carbon dioxide was not incorporated from the resultedinafunctional dismutase mechanism. Further optimi- overlaying atmosphere by aCO2 reduction pathway zations, with regard to the nature andstoichiometry of the ad- (Scheme 3). ditives, led to an efficient catalytic disproportionation (Supple- In an alternative route, the methanol formation could result mentaryTable 1–3). While initial attempts provided methanol from the directreduction of formaldehyde, either by another from paraformaldehyde in a1:1 stoichiometry,formalin decom- hydrogen-coupled processexploitingfree H2 as redox media- positionwith 1mol%of[Ru(p-cymene)Cl2]2 at 808Cinphos- tor (Scheme 2, centre), or by the dismutase-like dehydrogena- phate buffer (0.4m,pH6)proceeded with increased MeOH tive generation of reducing ruthenium-hydride species from yields (75%) as aresult of the formate dehydrogenation, which the tetrahedral formalin (Scheme 2, right). Examination of the allows for the reductionofasecond formaldehydeequiva- reactiongas phase by pressuremonitoring and headspace GC- lent.[35] For the primary disproportionation, variousmechanisms can be assumed(Scheme2). Hence, studies aiming to eluci- date the actual catalytic pathway have been conducted. Hydrolysis of paraformaldehyde would, in any of our propos- als, lead to free formaldehydeinequilibrium with its hydrated form methanediol. Initial investigations of the pH dependency of the reaction quickly revealed that the ruthenium-independ- ent Cannizzaro-type disproportionation appears only as aback- Scheme3.13C-Labelling experiment to exclude the possibility of carbon di- ground reaction at pH values greater than 9.5 (Supplementary oxideasformal disproportionation intermediate.

Chem. Eur.J.2016, 22,11568 –11573 www.chemeurj.org 11569 2016 Wiley-VCH Verlag GmbH &Co. KGaA, Weinheim Communication

TCD (gas chromatography with athermalconductivity detec- tor) analysis gave aclearevidence for the formation of hydro- gen gas in the initial period of the process. To our delight, we found that the formalin disproportionation was also taking place in an open flask setup rather than in an autoclave, in which strikingly high yields of methanolwere achieved even under constantremoval of the gas phase by argon purging

(Scheme 4a). Further confirmation of aH2-decoupled pathway

Figure 1. Crystallization of in situ formed formiato-bridged ruthenium dimers was achieved by precipitation as tetrafluoroborate salt. Crystal struc- ture analysisshowsasymmetric arrangementofthe cymene donors with both formate and chloride as bridging ligands. Although X-ray diffraction couldnot help to allocate the hydridic hydrogen between the two rutheni- um centres, clear evidence for the hydride nature of the complex was ob- tainedbyspectroscopicand spectrometric methods.[25] For detailed crystal- lographic data seethe SupportingInformation and ref. [35].

also revealed the rapid buildup of methanol, formate, and H2 duringthe first minutes. However,only the methanol produc- tion continued steadily over hours, explaining the good selec- tivity observed in the formalin decomposition (Figure 2).

Scheme4.Gas phase removal and deuterium labelling experiments to eluci- date the role of H2 gas as potentialredoxmediator. was obtainedfrom the results of the ruthenium-catalyzed dis- mutationofdeuterated formaldehyde (CD2O)n under H2 pres- sure. Therein, less than 2% of the partially hydrogen-contain- ing methanol(HCD2OD) were detected, which reflects exactly the isotopologiccompositionofthe commercial (CD2O)n used as startingmaterialinthis experiment (Supplementary Figure 3). The complete absence of H/D scrambling under the mixed H2/D2 atmosphere can serve as astrong endorsement for the disprovalofthe involvement of any gaseous redox me- Figure 2. Constant methanolformation versus discontinuous H2 release. diators (Scheme 4b). Hence, the ruthenium-catalysed formalin disproportionation is most likely to proceed by atruly dismu- tase-like mechanism with acatalyst-boundhydride mimicking While the underlying role of the phosphate additive still re- the intimate –nicotinamide arrangement. Acatalytically mains unclear and its elucidation will require amuch deeper potent dimerichydride complex has been isolated and charac- investigation, we expected that further fine-tuning of the reac- terized (Figure 1).[35] tion conditions would already allow us to provide ahighly effi- At first glance, these findings somewhat contradictedour cient formaldehyde-to-methanolprocedure. To also determine previousreport on the acceptorless H2 production from formal- the effect that modificationstothe ruthenium precatalyst in employing avery similar catalyst system. However,closein- impart upon the methanol yield of the reaction (Supplementa- spection of all parameters exposed the hidden key for this cat- ry table 7), arange of complexes analogoustothe commercial alytic bifurcation. As introduced during our attempt to exploit [Ru(p-cymene)Cl2]2 weresynthesized employing arecently de- the ruthenium-based dehydrogenase mimic in achemoenzy- veloped microwave-assisted protocol.[36] In the dismutation, matic approach,[34] the phosphate-containing reaction medium variation of the anionic counterion by different (pseudo)halides provedtohave amajor influence on the behavior of the di- showednoinfluence on the reaction rates or the methanol meric ruthenium catalyst. As opposed to the formalin dehydro- yield after 20 h(Table 1, Entries 1–4), which is in good agree- genationby[Ru(p-cymene)Cl2]2 in absence (or at low concen- ment with the previously described exchange of bridging li- trations) of phosphate, which led to astrong and constant rise gands during the methanediol dehydrogenation.[24] In contrast, in pressure,already 20 mol%ofK3PO4 substantially reduced the substitution pattern and particularly the polarity of the the initial hydrogen formation that quickly came to acomplete arene ligand provedtobedecisive for the activity of the meth- rest at approximately 5% conversion (Supplementary anol-generating process. While more hydrophobic h6 donors Figure 4,5). In situ analysis of the optimizeddismutation pro- led to decreased final concentrationsofthe alcohol (Table 1, cess by NMR spectroscopy and headspace GC-TCD analysis Entry 6), decoration of the arene by polar hydroxy groups, to

Chem. Eur.J.2016, 22,11568–11573 www.chemeurj.org 11570 2016 Wiley-VCH Verlag GmbH &Co. KGaA, Weinheim Communication

Table 1. Testing of the ruthenium complexesfor the self-sufficient meth- Table 2. Recharge experiments.[a] anol synthesis from paraformaldehyde.[a]

Entry [Ru(p-cymene)Cl ] CycleYield TON Entry Arene XTYield[b] 2 2 [mol%] [%] [8C] [%] 11175 50 1 p-cymene Cl 80 75 21275 100 2 p-cymene Br 80 68 31330 120 3 p-cymene I80 66 40.1 159 393 4 p-cymene SCN 80 68 50.1 212 473 5tolueneCl 80 70 60.1 33493 6hexamethylbenzene Cl 80 50 72-phenylethanolCl 80 73 [a] Reactions run on a4mmolscale in water (2 mL) in asealed autoclave. 82-phenoxyethanol Cl 80 79 Final concentrationsofmethanol were determined by 1HNMR spectros- 9 Cl 80 93 copy relativeto1,4-dioxane as internal standard. After each cycle, all vola- 10 Cl 25 58 tiles were removedinvacuo, and paraformaldehyde,water,and dioxane were replaced.

[a] Reactions run on a2mmol scale in aqueousphosphate buffer (1 mL, 0.4 m,pH6). [b] Conversions to methanol were determined by 1HNMR spectroscopyrelative to 1,4-dioxane as internal standard. All values are normalized to the reaction stoichiometry,only allowing amaximumof 67%methanol. facilitatethe aqueous solubility,resulted in slightly improved catalytic systems(Table1,Entries 7,8). By incorporation of acat- ionic 1,2-dimethylimidazolium unit,[37] ahighly active rutheni- um precatalyst was obtained that showedexcellent per- formance in the formalin dismutation with afinal methanol yield of 93 %(Table 1, Entry 9) along with deteriorated dehy- Figure 3. Recycling of the imidazolium-tagged arene–rutheniumdimer using drogenation properties (Supplementary Figure 6). More impor- nonwater-miscible ionic liquids as additives for the biphasic formaldehyde- tantly,the imidazolium-tagged dimer still exhibited agood ac- to-methanol conversion. tivity at considerably lower temperatures with areasonable conversion even at room temperature (Table 1, Entry 10). For this most reactive dismutase mimic, turnover frequencies up to portionation protocol.Initialattempts focusedonsecondary 1 1060 hÀ were recorded and asignificant methanol formation amines as additional reaction partners to study the redox self- was detectedatruthenium-loadings as low as 250 ppm (Sup- sufficient reductive amination employing formaldehyde both plementary Table 8). as carbon source and reducing agent.[38] To our delight, the Currently,there is asignificant focus on environmentally methylation of cyclic secondary amines proceeded smoothly in benign methods for chemical procedures and in particular on the presence of 0.5 mol %ofthe dimericcymene–ruthenium the use of recyclable catalysts. Initial recharge experimentsem- chloridewith excellent yields of N-methylpiperidine and N- ploying the commercialcymene–ruthenium dimer already re- methylmorpholine after only 2hof 81%and 69 %, respectively vealed astable catalytic systemwith considerable activity over (Scheme 5). Opening up an entirely new perspective on the at least two successive formaldehyde recharges(Table 2). In formalin redox chemistry,future investigationswill cover the subsequentcycles, the aqueous phase was replaced by an un- synthetic aspects of the formaldehyde decomposition in much treated formalin/phosphate solutionand 1-butyl-3-methylimi- greater detail. dazolium bis(trifluoromethanesulfonyl)imide ([bmim]NTf2)was In conclusion, we herein describe anew,homogeneously used as co-phase, in which asubstantially higher long-term catalyzed, selective formaldehyde-to-methanoltransformation stabilityofthe catalystwas achieved comparedtothe other as amissing piece in the C1-interconversion puzzle.Inspired by ionic liquids tested so far (Figure 3). Nonetheless, there remains bacterialformaldehyde dismutase biocatalystsusing arene– aneed for furtherinvestigations on the ionic liquid-based recy- ruthenium complexesasactivating species, it wasshown that cling systems, and in-depth studies to identify an optimal the reactionproceeds by aformaldehyde reduction by metal- methodology are currently ongoing. Based on the proposal that the methanol formation oc- curred throughreduction of free formaldehyde,the possibility of the in situ formation, and hydrogenations of even more re- active methylene compounds by condensation with the alde- hyde, appeared as valuable synthetic extensions of the dispro- Scheme5.Redoxself-sufficient reductive N-methylation.

Chem. Eur.J.2016, 22,11568 –11573 www.chemeurj.org 11571 2016 Wiley-VCH Verlag GmbH &Co. KGaA, Weinheim Communication bound hydrogen species as redox cofactor analogues,rather Haage-Stiftung (Max-Planck Society), and the Alexander-von- than through CO2 hydrogenation, formate decomposition, or Humboldt Foundation (Connect!program). Our special thanks the Cannizzaro disproportionation. An imidazolium-tagged go to SvetlanaBotov and Hagga Schmalz forthe access to the ruthenium complex did not only exhibit optimal catalytic prop- GC-TCD. erties, but further allowed for the construction of biphasic re- action setups that make use of ionic liquidsfor an easy catalyst Keywords: biocatalysis · ionic liquids · methanol · reductive recycling. In addition, the metal-mediated formalin decomposi- methylation · ruthenium tion could be employed in asynthetic manner,inwhich form- aldehydeacts as asole stoichiometric reagent in aredox self- sufficient reductivemethylation of secondary amines.

[1] W. Wang, S. Wang, X. Ma,J.Gong, Chem. Soc. Rev. 2011, 40,3703 –3727. [2] A. M. Bahmanpour,A.Hoadley,A.Tanksale, Rev.Chem.Eng. 2014, 30, Experimental Section 583–604. General remarks.All chemicals were used without further purifica- [3] B. Sam, B. Breit, M. J. Krische, Angew.Chem. Int. Ed. 2015, 54,3267 – tion, primarily from Sigma Aldrich or Strem Chemicals. Paraformal- 3274; Angew.Chem. 2015, 127,3317 –3325. [4] E. Alberico, M. Nielsen, Chem. Commun. 2015, 51,6714 –6725. dehyde was purchased from Acros Organics (>95 %Extra Pure), [5] S. Saeidi, N. A. S. Amin, M. R. Rahimpour, J. CO2 Util. 2014, 5,66–81. HPLC grade water was used as solvent without degassing. [6] S. Moret, P. J. Dyson,G.Laurenczy, Nat. Commun. 2014, 5,4017. 13 ( CH2O)n was obtained from the Cambridge Isotope Laboratories [7] R. Tanaka, M. Yamashita, K. Nozaki, J. Am. Chem.Soc. 2009, 131,14168 – 1 13 and (CD2O)n from Sigma Aldrich. Hand CNMR spectra were re- 14169. corded in CDCl or D O( 99.5%deuterated, purchased from Fluo- [8] K. Angermund, W. Baumann, E. Dinjus,R.Fornika, H. Gçrls, M. Kessler,C. 3 2  rochem) on aBruker Avance 300 (300 MHz). All chemical shifts (d) Krüger,W.Leitner,F.Lutz, Chem.Eur.J.1997, 3,755–764. are reported as parts per million (ppm) with reference to tetrame- [9] P. G. Jessop, Y. Hsiao, T. Ikariya, R. Noyori, J. Am. Chem. Soc. 1996, 118, thylsilane (TMS) (d=0.00 ppm) unless otherwise stated. Selectivi- 344–355. ties for the methanol production are given as percentage relative [10] P. Jessop,T.Ikariya, R. Noyori, Nature 1994, 368,231–233. [11] E. Graf, W. Leitner, J. Chem.Soc.,Chem. Commun. 1992,623–624. to the maximum paraformaldehyde conversion (ca. 67%) as an in- [12] C. A. Huff, M. S. Sanford, J. Am. Chem. Soc. 2011, 133,18122–18125. ternal standard and are the average of at least two runs unless oth- [13] C. A. Huff, M. Sanford, ACS Catal. 2013, 3,2412–2416. erwise mentioned. Headspace gas chromatography equipped with [14] E. Balaraman, C. Gunanathan, J. Zhang, L. J. W. Shimon, D. Milstein, Nat. aTCD was carried out on aThermo Fischer Scientific GC-TCD. Chem. 2011, 3,609–614. General procedure for the ruthenium-catalyzed formaldehyde [15] S. Wesselbaum, T. vom Stein, J. Klankermayer,W.Leitner, Angew.Chem. dismutation.Toaglassscrew capped reaction vial (20 mL) fur- Int. Ed. 2012, 51,7499 –7502; Angew.Chem. 2012, 124,7617 –7620. nished with astirrer,[Ru(1-phenethyl-2,3-dimethylimidazolium)Cl ] [16] S. Wesselbaum, V. Moha, M. Meuresch, S. Brosinski, K. M. Thenert, J. 3 2 Kothe, T. vom Stein, U. Englert, M. Hçlscher,J.Klankermayer,W.Leitner, (16 mg, 0.02 mmol) and paraformaldehyde (pFA)(60 mg, 2.0 mmol) Chem. Sci. 2015, 6,693–704. were added. K3PO4 (8 mg, 0.4 mmol) and the 1,4-dioxane standard [17] A. J. M. Miller, D. M. Heinekey,J.M.Mayer,K.I.Goldberg, Angew. Chem. were added, followed by HPLC grade water (2 mL). Afterwards, the Int. Ed. 2013, 52,3981 –3984; Angew.Chem. 2013, 125,4073 –4076. reaction vial was sealed and heated to the desired temperature for [18] T. J. Korstanje, J. I. van der Vlugt, C. J. Elsevier,B.deBruin, Science 2015, the respective reaction duration under constant stirring. Upon 350,298–320. completion, the reaction was cooled, shaken to ensure homogenei- [19] S. Savourey,G.Lef›vre, J.-C. Berthet,P.ThuØry, C. Genre, T. Cantat, ty,and an aliquot was taken directly for NMR analysis after dilution Angew.Chem. Int. Ed. 2014, 53,10466 –10470; Angew.Chem. 2014, 126, 10634–10638. with D2O. [20] H. Yurimoto,N.Kato, Y. Sakai, Chem. Rec. 2005, 5,367–375. General procedure for the ruthenium-catalyzed N-methylation. [21] S. Witthoff, K. Schmitz, S. Niedenführ,K.Nçh, S. Noack, M. Bott, J. Mar- Paraformaldehyde (300 mg, 10 mmol) and [Ru(p-cymene)Cl2]2 ienhagen, Appl. Environ. Microbiol. 2015, 81,2215 –2225. (6.1 mg, 0.01 mmol) were placed in ascrew-neck vial, followed by [22] A. Norin, P. W. van Ophem, S. R. Piersma,B.Persson,J.A.Duine, H. Jçrn- the addition of water (1 mL), and the secondary amine (piperidine, wall, Eur.J.Biochem. 1997, 248,282 –289. 170 mg, 2.0 mmol or morpholine (174 mg, 2.0 mmol). The vial was [23] J. Ras, P. W. van Ophem, W. N. Reijnders, R. J. van Spanning, J. A. Duine, closed, then placed in apreheated aluminium block (608C) and A. H. Stouthamer, N. Harms, J. Bacteriol. 1995, 177,247 –251. stirred for 2h.Afterwards, the vial was cooled to room tempera- [24] D. Tsuru, N. Oda,Y.Matsuo, S. Ishikawa,K.Ito, T. Yoshimoto, Biosci. Bio- ture, aqueous NaOH (2m,2mL) was added, and the aqueous technol.Biochem. 1997, 61,1354 –1357. [25] L. E. Heim, N. E. Schlçrer,J.-H.Choi, M. H. G. Prechtl, Nat. Commun. phase was washed with DCM (3”10mL). The combined organic 2014, 5,3621. fractions were dried over MgSO4 and Al2O3,and the solvent was re- [26] N. Kato,K.Shirakawa, H. Kobayashi, C. Sakazawa, Agric.Biol. Chem. moved in vacuo to yield the desired amine (N-methylpiperidine, 1983, 47,39–46. 160 mg, 1.62 mmol, 81 %; or N-methylmorpholine, 140 mg, [27] L. V. Bystrykh, N. I. Govorukhina, P. W. van Ophem, H. J. Hektor,L.Dij- 1.38 mmol, 69%). khuizen,J.A.Duine, J. Gen. Microbiol. 1993, 139,1979–1985. [28] H. Yanase, H. Noda, K. Aoki, K. Kita, N. Kato, Biosci. Biotechnol. Biochem. 1995, 59,197–202. Acknowledgements [29] Comparison of sequencesobtained from the RCSB protein data bank (PDB) data for formaldehyde dehydrogenase(1KOL): a) N. Tanaka, Y. Ku- sakabe, K. Ito, T. Yoshimoto, K. T. Nakamura, J. Mol. Biol. 2002, 324,519 – We gratefully acknowledge financial support provided by the 533;and formaldehyde dismutase (2DPH): b) T. Hasegawa, A. Yamano, Ministerium für Innovation, Wissenschaft und Forschung (NRW- K. Miura, Y. Katsube, H. Yanase, N. Kato, Acta Crystallogr.Sect. A 2002, returnee award 2009toM.H.G.P.), the Deutsche Forschungsge- 58,c102. meinschaft (Heisenberg-Program), the COST (Catalytic Rou- [30] N. Kato, T. Yamagami, M. Shimao,C.Sakazawa, Eur.J.Biochem. 1986, 156,59–64. tines for Small Molecule Activation, CARISMA), the Ernst- [31] R. P. Mason, J. K. M. Sanders, Biochemistry 1989, 28,2160 –2168.

Chem. Eur.J.2016, 22,11568 –11573 www.chemeurj.org 11572 2016 Wiley-VCH Verlag GmbH &Co. KGaA, Weinheim Communication

[32] For the catalytic disproportionation of higheraldehydes, see:M.Bian- paper.These data are provided free of chargebyThe Cambridge Crys- chi, U. Matteoli, G. Menchi, P. Frediani,F.Piacenti, J. Organomet. Chem. tallographic Data Centre 1982, 240,65–70. [36] L. E. Heim,S.Vallazza, D. van der Waals, M. H. G. Prechtl, Green Chem. [33] For aruthenium-catalyzed Tishchenko-type disproportionation, see:M.- 2016, 18,1469 –1474. O. Simon, S. Darses, Adv.Synth. Catal. 2009, 351,153. [37] T. J. Geldbach, P. J. Dyson, J. Am.Chem. Soc. 2004, 126,8114–8115. [34] L. E. Heim,D.Thiel, C. Gedig,J.Deska, M. H. G. Prechtl, Angew.Chem. [38] For aformate-dependent methylation of anilines, see:S.Savourey, G. Int. Ed. 2015, 54,10308–10312; Angew.Chem. 2015, 127,10447 –10451. Lef›vre, J.-C. Berthet,T.Cantat, Chem. Commun. 2014, 50,14033–

[35] As aresult of the 2:1stoichiometry of the reaction (1 equiv of CH2O 14036.

giving 0.67equiv MeOH and 0.33equiv of CO2,the yields reported herein are normalized to the maximum outcome of 0.67 equiv). For the crystal structureofthe synthesized ruthenium complex,see: CCDC Received:June 6, 2016 1471617 contain the supplementary crystallographic data for this Published online on July 6, 2016

Chem. Eur.J.2016, 22,11568 –11573 www.chemeurj.org 11573 2016 Wiley-VCH Verlag GmbH &Co. KGaA, Weinheim