Metal–ligand cooperation behaviour of Fe and Co complexes bearing a tetradentate phenanthroline-based PNNP ligand

Yumiko Nakajima *ab, Tomohiro Takeshita b and Nai-Yuan Jheng ab
aInterdisciplinary Research Center for Catalytic Chemistry (IRC3), National Institute of Advanced Industrial Science and Technology (AIST), 1-1-1 Higashi, Tsukuba, Ibaraki 305-8565, Japan. E-mail: yumiko-nakajima@aist.go.jp
bGraduate School of Pure and Applied Sciences, University of Tsukuba, 1-1-1 Tennoudai, Tsukuba, Ibaraki 305-8577, Japan

Received 12th February 2021 , Accepted 12th March 2021

First published on 31st March 2021


This perspective article describes the synthesis of a series of Fe and Co complexes coordinated with a phenanthroline-based meridional PNNP ligand (2,9-bis((diphenylphosphino)methyl)-1,10-phenanthroline). PNNP–iron(II) dichloride and –cobalt(I) chloride, [FeCl2(PNNP)] and [CoCl(PNNP)], underwent abstraction of the benzylic H-atom upon treatment with NaOtBu, forming the corresponding deprotonated products [FeCl(PNNP′)] (1) and [Co(PNNP′)] (2), respectively, each of which bears an asymmetrical PNNP′ ligand with a dearomatized phenanthroline backbone as a good metal–ligand cooperation (MLC) scaffold. Complex 2 achieved facile H–H bond cleavage mediated by unique long-range MLC, where the PNNP backbone acts as a H-atom reservoir.


Introduction

3d Metal complexes have recently drawn increased attention due to their potential capability to act as cheap and non-toxic surrogates for precious-metal catalysts.1 It is also known that 3d metal complexes exhibit unique reactivity such as one-electron (1e) redox behaviour. Owing to these properties, 3d metal complexes have recently enhanced their significance in catalysis. However, examples of their well-defined reaction chemistry still remain scarce compared with those of 4d and 5d metal complexes, mainly due to the difficulty in controlling the diverse and complicated electronic states of 3d metal complexes.

In this context, we have thus far focused on Fe and Co complexes bearing a phenanthroline-based tetradentate P–N–N–P ligand, 2,9-bis((diphenylphosphino)methyl)-1,10-phenanthroline (PNNP).2,3 So far, various kinds of P–N–N–P ligand systems were developed and utilised in a series of catalytic reactions. For example, Morris et al. demonstrated the first well-defined Fe-based catalyst bearing a P–N–N–P ligand with phosphorus sidearms and an imine- and/or amine-backbone for asymmetric hydrogenation (AH) and transfer hydrogenation (ATH).4 Saito et al. reported that a bipyridine-based P–N–N–P ligand, 6,6′-bis((dialkylphosphino)methyl)-2,2′-bipyridine, offers “sterically confined bipyridine–metal framework”, and their ruthenium complexes exhibit good catalytic activity towards hydrogenolysis of inactivated amides.2c In both reaction systems, the bifunctional combination of a metal–hydride bond and a nitrogen–hydrogen bond of the ligand backbone plays a pivotal role during the reactions. Milstein recently reported that a Ru complex bearing a phenanthroline-based P–N–N–P ligand, 2,9-bis((di-tert-butylphosphino)methyl)-1,10-phenanthroline, exhibits unique long-range metal–ligand cooperation (MLC), where the hydride migrates to the endocyclic phenanthroline backbone.5

The PNNP ligand of interest in this research connects with a metal centre through a rigid meridional coordination mode with four σ-donating moieties, and thus causes large crystal field splitting to the metal centre. Because of these electronic properties, various PNNP-supported Fe and Co complexes indeed preferably act as stable diamagnetic complexes with a low spin state. As a result, the precise design of PNNP–Fe and –Co complexes was easily achieved, and furthermore, we successfully applied the MLC concept.

The MLC concept has brought significant breakthroughs in, mainly, 4d and 5d metal catalysis so far.6 One important landmark has been made by Milstein's group using acridine- and pyridine-based PNP and PNN pincer systems.7 In their systems, various bond cleavages proceed, which is associated with an aromatization–dearomatization sequence of the ligand backbone, enabling us to achieve various catalytic reactions (Scheme 1).


image file: d1dt00476j-s1.tif
Scheme 1 Bond cleavage mediated via MLC of pyridine-based pincer systems.

Studies on the MLC of pincer-3d metal complexes have appeared quite recently, e.g., in the late 2010s.8 However, there is still room for development in this area considering the great advances made in the MLC of pincer-4d and -5d metal complexes.

This perspective article describes the synthesis of PNNP–Fe and –Co complexes 1 and 2, which possess a dearomatized ligand backbone with an exo-methylene carbon as a typical useful scaffold for bond cleavage via MLC (Fig. 1).3 In the PNNP–Co system, peculiar H2 uptake was achieved, which is mediated via long-range MLC.


image file: d1dt00476j-f1.tif
Fig. 1 Structures of 1 and 2.

Results and discussion

Synthesis of PNNP-R-supported Fe and Co complexes

PNNP-supported iron(II) dichloride [FeCl2(PNNP)] (3) was quantitatively synthesized by the reaction of FeCl2 with PNNP in THF under reflux (Scheme 2).3a Complexation of a Co(I) precursor, [CoCl(PPh3)3], with PNNP proceeded at ambient temperature to form [CoCl(PNNP)] (4) as a sole product.3d
image file: d1dt00476j-s2.tif
Scheme 2 Synthesis of 3 and 4.

Next, deprotonation of 3 and 4 was performed by treating them with NaOtBu. The reaction of 3 with NaOtBu (2 equiv.) at −78 °C resulted in the formation of 1, which is supported by an asymmetrical PNNP′ ligand with a dearomatized phenanthroline backbone, as a major product. A trace amount of [FeOtBu(PNNP′)] (5), as well as several unidentified minor compounds, was also formed in this reaction (Scheme 3).3b


image file: d1dt00476j-s3.tif
Scheme 3 Deprotonation of 3 with NaOtBu.

It was also confirmed that selective formation of 1 was achieved by the reaction of FeCl2 with a PNNP′-coordinated Li complex [Li(OEt)(PNNP′)], which was alternatively prepared from PNNP and nBuLi in Et2O (Scheme 4).3b


image file: d1dt00476j-s4.tif
Scheme 4 Selective synthesis of 1 using a PNNP′–Li complex.

The reaction of 4 with NaOtBu (1 equiv.) similarly proceeded at room temperature to form the corresponding deprotonated complex [Co(PNNP′)] (2) as a sole product (Scheme 5).3d Complexes 1 and 2 are stable diamagnetic compounds and were fully identified using NMR spectroscopy. The asymmetric dearomatized structure of the PNNP′ ligand in 1, 2, and 5 was confirmed by single-crystal X-ray analysis. The C1–P1 bonds are 1.777(4) Å for 1, 1.7882(2) Å for 2 and 1.756(4) for 5, which are significantly shorter than the C2–P2 bond lengths with a typical single C–P bond (1.84–1.87 Å).9


image file: d1dt00476j-s5.tif
Scheme 5 Deprotonation of 4 by NaOtBu.

This could be explained by considering the resonance structures of the metallacycle moiety composed of P1, C1, C3, N1 and metal (Fig. 2). Such rather short P–C bonds are also observed in the previously reported dearomatized-PNNP, -PNN, and -PNP systems.10


image file: d1dt00476j-f2.tif
Fig. 2 Resonance structures of 1 and 2.

Metal–ligand cooperation of PNNP–Fe and –Co complexes

With a good structural motif of the dearomatized-PNNP′ ligand, we next investigated the reactivities of 1 and 2 towards bond cleavage. We found that 1 did not react with H2, H2O or alcohols.3b In contrast, complex 2 easily cleaved the H–H bond at ambient temperature to form [Co(PNNP′′)] (6) (Scheme 6).3d To our surprise, two H termini, formed during the reaction, were incorporated into the endocyclic double bond of the ligand backbone. As a result, 6 exhibits a four-coordinate structure bearing a new asymmetric PNNP′′ ligand. A further interesting point is that 6 partially converted to 2 (24% NMR yield) upon heating at 110 °C for 16 h. Such unique H atom uptake on the PNNP backbone of the phenanthroline-based P–N–N–P ligand was also previously reported as long-range MLC by Milstein's group using the Ru system, where the phenanthroline-backbone plays the role of a H atom acceptor.5 In this system, the H atoms were initially provided by a hydride reagent or alcohol. On the other hand, it should be mentioned that the reversible uptake of molecular hydrogen via facile H–H cleavage was achieved in our PNNP–Co MLC system. Although mechanistic details of the unique MLC behaviour were not well-disclosed in the previous Ru-system, we assumed that the formation process of 6 includes the hydride intermediate A, which could be formed via H–H cleavage by the conventional MLC (Scheme 6, below).
image file: d1dt00476j-s6.tif
Scheme 6 H–H cleavage by 2.

Different from the above-mentioned study, we recently undertook some experiments in an attempt to isolate reactive Co(I) alkyl complexes and revealed their unique 1e redox behaviour, which helps to understand the formation process of 6 from 2.

Cobalt alkyl complexes [Co(R)(PNNP)] (7: R = Me; 8: R = CH2SiMe3) were synthesized by the reaction of 1 with the corresponding Grignard reagent. Although 7 was not stable at ambient temperature and gradually decomposed to form several unidentified compounds, 8 was stable enough to be isolated as a pure form in 87% yield. It is known that low-coordinate Co(I) species are reactive towards C–H bond cleavage.11 For example, Chirik's group reported a facile C–H bond cleavage of benzene using [CoMe(PNP)] at 80 °C.11a In contrast, the heating C6D6 solution of 8 resulted in the clean formation of 2 (86% yield), which was accompanied by the formation of SiMe4 (Scheme 7).3d


image file: d1dt00476j-s7.tif
Scheme 7 Transformation of 8 into 2.

Considering the facile radical behaviour of Co complexes,12 we initially hypothesized that the observed unique structural transformation from 8 to 2 is triggered by Co–CH2SiMe3 homolysis, and successive benzylic H-atom abstraction by the formed alkyl radical, ˙CH2SiMe3. To elucidate this point, we next performed the reaction of 8 with excess TEMPO ((2,2,6,6-tetramethylpiperidin-1-yl)oxyl) as a radical scavenger (Scheme 8). As expected, TEMPO–CH2SiMe3 was formed after the reaction, however as a minor by-product, and TEMPO–H was mainly obtained (TEMPO–H/TEMPO–CH2SiMe3 = 7/1). Thus, the reaction might also proceed by the initial benzylic C–H homolysis and subsequent CH2SiMe3 abstraction to form SiMe4.13 In this reaction, 2 was not formed, but the reaction resulted in the formation of a Co(III)–alkyl complex [Co(CH2SiMe3)(PNNP′′′)] (9) as a major product, which should be formed via facile benzylic H-atom abstraction by TEMPO.


image file: d1dt00476j-s8.tif
Scheme 8 Reaction of 8 with TEMPO.

The above-mentioned experimental results suggest that the possible hydride intermediate A could be transferred to 2 initiated by homolysis of either a Co–H bond or a benzylic C–H bond. To further investigate the mechanism of the unique long-range MLC, we recently synthesized a PNNP–Co ethyl complex as a potential PNNP–Co hydride precursor since all the attempts to synthesise hydride intermediates using hydride reagents were not successful. The result will be reported in due course.

Conclusions

The manuscript summarises the synthesis of a series of PNNP–Fe and –Co complexes. Thanks to the strong σ-donating ability of the PNNP ligand, PNNP-coordinated complexes behave as stable diamagnetic species, and thus a concept of MLC was expanded to both PNNP–Fe and –Co systems based on the precise design of the reaction space to demonstrate unique long-range MLC; i.e. reversible H–H bond cleavage was achieved, where the phenanthroline backbone behaves as a molecular hydrogen reservoir. The mechanistic study supported that the observed MLC behaviour of the PNNP–Co systems is based on the radical properties of Co complexes. Specifically, inclusion of the occurrence of Co–H homolysis and/or benzylic C–H homolysis was demonstrated based on the mechanistic study using PNNP–Co alkyl analogues. Studies to reveal the potential applicability of PNNP-3d complexes as hydrogenolysis catalysts of various organic substrates are now underway in our laboratory. We believe that our PNNP–Fe and –Co systems will offer a new tool towards achieving characteristic base-metal-catalysed transformation reactions.

Author contributions

N.-Y. J. and T. T. conducted the experiments and wrote the manuscript. Y. N. directed the research and wrote the manuscript.

Conflicts of interest

There are no conflicts to declare.

Acknowledgements

This work was supported by the MEXT Grant-in-Aid for Scientific Research on Innovative Areas “Precise Formation of a Catalyst Having a Specified Field for Use in Extremely Difficult Substrate Conversion Reactions” (no. 18H04280) and by KAKENHI (no. 18H01986) from JSPS.

Notes and references

  1. (a) B. Plietker, Iron Catalysis in Organic Chemistry, Wiley-VCH, Weinheim, Germany, 2008 Search PubMed ; (b) I. Bauer and H.-J. Knölker, Chem. Rev., 2015, 115, 3170 Search PubMed ; (c) C. Bolm, J. Legros, J. Le Paih and L. Zani, Chem. Rev., 2004, 104, 6217 Search PubMed ; (d) H. Pellissier and H. Clavier, Chem. Rev., 2014, 114, 2775 Search PubMed ; (e) A. Fürstner, ACS Cent. Sci., 2016, 2, 778 Search PubMed .
  2. (a) R. Ziessel, Tetrahedron Lett., 1989, 30, 463 Search PubMed ; (b) T. Miura, M. Naruto, K. Toda, T. Shimomura and S. Saito, Sci. Rep., 2017, 7, 1586 Search PubMed ; (c) S. Saito, N. Ryoji, T. Miura, M. Naruto, K. Iida, Y. Takada, K. Toda, S. Nimura, S. Agrawal and S. Lee, US Pat0107151A1, 2016 Search PubMed .
  3. (a) T. Takeshita, K. Sato and Y. Nakajima, Dalton Trans., 2018, 47, 17004 Search PubMed ; (b) T. Takeshita and Y. Nakajima, Chem. Lett., 2019, 48, 364 Search PubMed ; (c) M. Gautam, T. Yatabe, S. Tanaka, N. Satou, T. Takeshita, K. Yamaguchi and Y. Nakajima, ChemistrySelect, 2020, 5, 15 Search PubMed ; (d) N.-Y. Jheng, Y. Ishizaka, Y. Naganawa, A. Sekiguchi and Y. Nakajima, Dalton Trans., 2020, 49, 14592 Search PubMed .
  4. (a) R. H. Morris, Chem. Soc. Rev., 2009, 38, 2282 Search PubMed ; (b) R. H. Morris, Acc. Chem. Res., 2015, 48, 1494 Search PubMed .
  5. R. Langer, I. Fuchs, M. Vogt, E. Balaraman, Y. Diskin-Posner, L. J. W. Shimon, Y. Ben-David and D. Milstein, Chem. – Eur. J., 2013, 19, 3407 Search PubMed .
  6. (a) J. R. Khusnutdinova and D. Milstein, Angew. Chem., Int. Ed., 2015, 54, 12236 Search PubMed ; (b) H. Grützmacher, Angew. Chem., Int. Ed., 2008, 47, 1814 Search PubMed ; (c) T. Higashi, S. Kusumoto and K. Nozaki, Chem. Rev., 2019, 119, 10393 Search PubMed .
  7. (a) J. Zhang, G. Leitus, Y. Ben-David and D. Milstein, J. Am. Chem. Soc., 2005, 127, 10840 Search PubMed ; (b) C. Gunanathan and D. Milstein, Chem. Rev., 2014, 114, 12024 Search PubMed .
  8. (a) G. Bauer and X. Hu, Inorg. Chem. Front., 2016, 3, 741 Search PubMed ; (b) N. Gorgas and K. Kirchner, Acc. Chem. Res., 2018, 51, 1558 Search PubMed ; (c) W. Lu, B. Sahoo, K. Junge and M. Beller, Acc. Chem. Res., 2018, 51, 1558 Search PubMed ; (d) L. Alig, M. Fritz and S. Schneider, Chem. Rev., 2019, 119, 2681 Search PubMed .
  9. The sum of covalent radii of the C and P atoms is 1.87 Å; J. A. Dean, Lange's handbook of chemistry, MacGrawHill, Inc., New York, St. Louis, San Francisco, Auckland, Bogotä, Caracas, Lisbon, London, Madrid, Mexico, Milan, Montreal, New Delhi, Paris, San Juan, São Paulo, Singapore, Sydney, Tokyo, Tronto, 15th edn, 1999 Search PubMed .
  10. (a) H. Qichen, X. Minzhi, Q. Yanlong and X. Weihua, J. Organomet. Chem., 1985, 287, 419 Search PubMed ; (b) A. Sacco, G. Vasapollo, C. F. Nobile, A. Piergiovanni, M. A. Pellinghelli and M. Lanfranchi, J. Organomet. Chem., 1988, 356, 397 Search PubMed ; (c) M. Feller, E. Ben-Ari, M. A. Iron, Y. Diskin-Posner, G. Leitus, L. J. W. Shimon, L. Konstantinovski and D. Milstein, Inorg. Chem., 2010, 49, 1615 Search PubMed ; (d) Y. Nakajima and F. Ozawa, Organometallics, 2012, 31, 2009 Search PubMed .
  11. (a) S. P. Semproni, C. C. Hojilla Atoenza and P. J. Chirik, Chem. Sci., 2014, 5, 1956 Search PubMed ; (b) F. Hung-Low, J. P. Krogman, J. W. Tye and C. A. Bradley, Chem. Commun., 2012, 48, 368 Search PubMed ; (c) S. C. Marinescu, J. R. Winkler and H. B. Gray, Proc. Natl. Acad. Sci. U. S. A., 2012, 109, 15127 Search PubMed ; (d) O. M. Reinaud and K. H. Theopold, J. Am. Chem. Soc., 1994, 116, 6979 Search PubMed ; (e) S. Sung, Q. Wang, T. Krämer and R. D. Young, Chem. Sci., 2018, 9, 8234 Search PubMed .
  12. (a) C. Y. Mok and J. F. Endicott, J. Am. Chem. Soc., 1977, 99, 1276 Search PubMed ; (b) A. M. Stolzenberg and Y. Cao, J. Am. Chem. Soc., 2001, 123, 9078 Search PubMed ; (c) R. P. Yu, J. M. Darmon, C. Milsmann, G. W. Margulieux, S. C. E. Stieber, S. DeBeer and P. J. Chirik, J. Am. Chem. Soc., 2013, 135, 13168 Search PubMed ; (d) M. Giedyk, K. Goliszewska and D. Gryko, Chem. Soc. Rev., 2015, 44, 3391 Search PubMed ; (e) G. Li, J. L. Kuo, A. Han, J. M. Abuyuan, L. C. Young, J. R. Norton and J. H. Palmer, J. Am. Chem. Soc., 2016, 138, 7698 Search PubMed ; (f) K. Imabeppu, H. Kuwano, E. Yutani, H. Kitagishi and K. Kano, Eur. J. Inorg. Chem., 2016, 2016, 1784 Search PubMed .
  13. Related to this reaction path, a benzylic C–H bond homolysis is also reported for the PNP–Co(I) methyl complex; E. Khaskin, Y. Diskin-Posner, L. Weiner, G. Leitus and D. Milstein, Chem. Commun., 2013, 49, 2771 Search PubMed  . However, the reaction path that is initiated by the reduction of TEMPO and successive deprotonation at the benzylic position by the in situ formed TEMPO(–) cannot be completely excluded.

This journal is © The Royal Society of Chemistry 2021