Iridium-Catalyzed Regio- and Diastereoselective Synthesis of C-Substituted Piperazines

Piperazine rings are essential motifs frequently found in commercial drugs. However, synthetic methodologies are mainly limited to N-substituted piperazines, preventing structural diversity. Disclosed herein is a straightforward catalytic method for the synthesis of complex C-substituted piperazines based on an uncommon head-to-head coupling of easily prepared imines. This 100% atom-economic process allows the selective formation of a sole diastereoisomer, a broad substrate scope, and a good functional group tolerance employing a bench-stable iridium catalyst under mild reaction conditions. Key to the success is the addition of N-oxides to the reaction mixture, as they notably enhance the catalytic activity and selectivity.


Starting materials and physical methods
The operations were performed under an argon atmosphere using standard Schlenk techniques and glovebox facilities. Aldehydes and amines were were adquired commercially and distilled before use. C 6 D 6 and CD 3 CN were dried over 4 Å molecular sieves and degassed through three freeze-pump-thaw cycles. [{Ir(µ-Cl)(cod)} 2 ] (1) was prepared according to the literature method. S1 All other reagents and solvents were adquired commercially and used as received unless otherwise stated. Carbon, hydrogen, and nitrogen analyses were carried out with a Perkin-Elmer 2400 CHNS/O microanalyzer. NMR spectra were recorded on Bruker AV300, AV400 and AV500 spectrometers operating at 300.13, 400.13 MHz and 500.13 MHz, respectively, for 1 H. Chemical shifts are reported in ppm and referenced to SiMe 4 , using the internal signal of the deuterated solvent ( 1 H and 13 C). 31 P NMR chemical shifts are reported relative to external 85 % H 3 PO 4 . 1 H NMR spectra for quantitative measurements were recorded using the standard sequence from Bruker zg30, with 8 scans and a delay (d1) of 15 s, which ensures a good relative intensity of the selected resonances of complexes/substrates/products and that from the internal standard (Me signal of toluene). Mass spectra and high resolution mass specta of complexes were acquired on a Bruker Esquire3000 plus (ESI+) and a Bruker MicroTOF-Q (ESI+) spectrometers, respectively. Conductivities were measured in methanol solutions using a Philips PW 9501/01 conductivity meter.

Synthesis of complexes and monitoring of reactions [IrCl(cod)(PPh 3 )] (5):
was prepared according to the modified literature method as described. S2 [{Ir(µ-Cl)(cod)} 2 ] (1, 542.0 mg, 0.806 mmol) and PPh 3 (423.3 mg, 1.612 mmol) were dissolved in toluene (12 mL) and the reaction was stirred for 15 minutes yielding an orange suspension. The suspension was filtered throught celite, and the filtrate was concentrated under reduced pressure to approx. 4 mL and layered with n-hexane (10 mL). After 3 days, the product was isolated as orange crystals, which were separated by decantation, washed with hexane (2 x 5 mL) and vacuum-dried. Yield: 395.3 mg (82%).  was added to an NMR tube containing complex 5 (10.0 mg, 0.0167 mmol) in C 6 D 6 (0.5 mL). The reaction was monitored by 1 H NMR spectroscopy observing the quantitative conversion of the imine to the piperazine 6a after 6 h at rt ( Figure S1, b).

Synthesis of imines
General procedure for the synthesis of imines (for imines 2a-2c, 2e, 2g-2j): Freshly distilled equimolar amounts of the aldehyde (0.01 mol) and amine (0.01 mol) were added to a suspension of MgSO 4 (5.0 g, 0.0415 mol) in dichloromethane (10 mL). The reaction mixture was stirred for 3 h in the absence of light, filtered through celite and the filtrate was vacuum-dried. The imines were stored in a −20 ºC freezer until use and the 1 H NMR spectrum was checked before use. Imines 2a, S3 2g, S4 2e, S4 2l, S5 and 2k S6 were identified by comparison to the data reported in literature.
Imines 2d S4 , 2f S7 and 2m S8 were prepared according to the procedure reported in the literature.
Unreported imines (or those whose NMR data is unreported, 2b, 2c, 2h, 2i, 2j) were characterized by NMR spectroscopy and mass spectrometry and the information is detailed below. For NMR spectra of these imines see Figures S10-S19.

Catalytic experiments and synthesis of piperazines
Standard procedure for the synthesis of 6a with an additive. In a glovebox, complex 5 (5.0 mg, 0.0084 mmol) and the corresponding additive (0.084 mmol) were weighed into an NMR tube and dissolved in the appropriate amount of C 6 D 6 to get a final volume of 0.5 mL. Then, the internal standard (toluene, 8.0 µL, 0.075mmol) and 2a (74 µL, 0.418 mmol) were added. The NMR tube was sealed and shaken, and at this point the chronometer was switched on. The tube was taken out of the glovebox and immediately loaded into the NMR spectrometer (around 3-5 min). 1 H NMR spectra of the mixture were recorded at different intervals of time. The time at which each spectrum was acquired was directly taken from the Mestre-software (view/table/parameters). See Table 1 and Figure 3 in the main text for details.
Following a similar procedure, the influence of the solvent was analyzed in selected cases, and the results are depicted in Figure S2. Using Me 3 NO•2H 2 O as the additive, a faster reaction in acetonitrile than in benzene was observed, probably due to a better solubility of the amine oxide in acetonitrile. However, in the presence of NEt 3 , a non-polar solvent such as benzene produced the best results. Standard procedure for the synthesis of tetrasubstituted piperazines (6a-6i) with Me 3 NO•2H 2 O. In a glovebox, complex 5 (5.0 mg, 0.0084 mmol) and Me 3 NO•2H 2 O (9.3 mg, 0.084 mmol) were weighed into an NMR tube and dissolved in the appropriate amount of C 6 D 6 to get a final volume of 0.5 mL. Then, the internal standard (toluene, 8.0 µL), and the imine (0.418 mmol) were added. The NMR tube was sealed and shaken. At this point, the chronometer was switched on, the tube was taken out of the glovebox and immediately loaded into the NMR spectrometer (around 3-5 min). 1 H NMR spectra were recorded at different intervals of time; the time at which each spectrum was acquired was directly taken from the Mestre-software (view/table/parameters). All reactions were monitored by NMR spectroscopy ( Figure S3). After that, the solutions were filtered through an alumina plug and washed with chloroform (3 mL). The filtrate was vacuum-dried to yield the piperazines, isolated generally as oils. For NMR spectra of the piperazines see Figures S20-S37. In the absence of catalyst, no conversion to the corresponding piperazines was observed either in the presence or absence of Me 3 NO (entries 1-4, Table S1). In the particular case of imines 2h and 2i, they converted to the corresponding piperazines in the absence of Me 3 NO with comparable reaction times (entries 5-8, Table S1).

X-ray diffraction studies on complex [Ir(cod)(C 24 H 22 N 6 )]Cl [4]Cl and piperazines 6a, 6d, and 6i
Intensity measurements were collected with a Siemens Smart Apex (6a and 6i) or a Bruker D8 Venture ([4]Cl and 6d) diffractometers, with Mo Kα radiation at 100 K. A semi-empirical absorption correction was applied to the data sets with the multi-scan [S9] methods. The structures were solved by direct methods with SHELX-97 (6a and 6i) or SHELXT-2014 [S10] ([4]Cl and 6d) and refined by full-matrix least-squares on F 2 with the program SHELXL-2016, [S11] in the WINGX [S12] package. All non-hydrogen atoms were refined with anisotropic displacement parameters, except in the disordered part of [4]Cl. In the models of 6a and 6d, the hydrogen atoms were located in difference-Fourier maps and refined free, including the isotropic displacement parameters. In 6i the hydrogen atoms were geometrically calculated and refined by the riding mode including the isotropic displacement parameters, except the hydrogens bonded to nitrogen atoms that were located in a difference-Fourier map and refined by the riding mode, but with free isotropic displacement parameters. In compound [4]Cl, the procedure SQUEEZE [S13] was used to model a severe solvent disorder, accounting for one hexane molecule per unit cell. Besides that, several atoms in the main molecule also are disordered; these disordered atoms were refined with isotropic displacement parameters and geometrical constraints, and the corresponding hydrogen atoms were not included in the model. The remaining hydrogen atoms were geometrically calculated and refined by the riding mode including the isotropic displacement parameters.                           H}-apt NMR spectrum of 2,3-diisopropyl-5,6-di(pyridin-2-yl)piperazine (6h) in CDCl 3 . The asterisk (*) indicates the residual signal of the solvent.  H}-apt NMR spectrum of 2,3-diisobutyl-5,6-di(pyridin-2-yl)piperazine (6i) in CDCl 3 . The asterisk (*) indicates the residual signal of the solvent.