Covalent self-assembly of the speci fi c RSSR isomer of 14-membered tetrakisphosphine †

Macrocyclic polyphosphines exist as a rule as mixtures of diastereomers due to the high inversion barrier of phosphorus atoms in phosphines (150 kJ mol). The separation of these mixtures and isolation of individual isomers in a pure state is a serious obstacle for the use of macrocyclic phosphines in coordination and supramolecular chemistry, as well as in catalysis. We have proposed the stereoselective synthesis of macrocyclic tetrakisphosphines via the covalent self-assembly approach by the condensation of 1,n-bisphosphinealkanes (n = 2–5) with formaldehyde and primary amines. This method of the design of macrocyclic polyphosphines was successfully used for the stereoselective synthesis of more than forty representatives of 14-, 16-, 18and 20-membered macrocyclic P4N2 corands. According to the Le Bel–van’t Hoff rule, the maximum number of isomers for four asymmetric atoms is 2, whereas the number of diastereoisomers for tetrakisphosphines is five. Nevertheless instead of five possible diastereomers only one stereoisomer of 14-, 16-, 18and 20-P4N2 was isolated from reaction mixtures in good or moderate yield. The crystal structure analysis of the at least thirty isolated P4N2 corands allowed us to formulate the empirical rule which predicted the configuration of the phosphorus atoms in the macrocyclic tetrakisphosphine corands. So, if the two chiral phosphorus centers in the macrocycle are linked by an odd number of methylene groups, then the RPSPSPRP stereoisomer is adopted, that is realized for 16and 20-membered corands. If the phosphorus atoms in the macrocycle are linked by an aliphatic chain consisting of an even number of methylene groups, then the racemic SPSPSPSP/RPRPRPRP isomer is formed. The selectivity of the proposed synthetic approach may be explained by covalent self-assembly phenomena. The distinctive feature of the covalent self-assembly processes is their ability of self-correction when the “incorrect” intermediate is able to decompose into starting compounds due to the reversibility of the reaction. These compounds react further to give a more thermodynamically stable “correct” product. The reaction mixtures of secondary bisphosphines with formaldehyde and primary amines contain many products but only one of them is crystallized at the final stage. It should be mentioned that in contrast to the 16-, 18 and 20-membered macrocycles, which are formed from 1,n(bisphosphine)propane, butane and pentane either 1-aza-3,6diphosphacycloheptanes or 14-membered tetrakisphosphines are formed in the course of interaction of 1,2-bis (phenylphosphino)ethane, formaldehyde and primary alkyland alkylarylamine. It was shown that dissolving of 14-P4N2 macrocycles led to the splitting of the macrocycle into the RSand RR/SS isomers of 1-aza-3,6-diphosphacycloheptane. The higher homologues of P4N2-macrocycles are only stereoisomerized in solutions to some extent. The lability of P–CH2–N fragments in cyclic and macrocyclic aminomethylphosphines is responsible for the reversibility of the condensation reaction whereas crystallization is a driving force of the stereoselective formation of aminomethylphosphine macrocycles. Despite the presence of different isomers of P4N2 corands in the reaction mixtures all attempts to isolate the isomers which are the exception to the empirical rule were unsuccessful. Here we represent the first example of the unexpected RSSR isomer of 14-membered macrocyclic aminomethylphosphine and its dicopper complex. The interaction of 1,2-bis (phenylphosphino)ethane (1) with formaldehyde and iso†Electronic supplementary information (ESI) available: Synthetic procedures, NMR-experiments (Fig. S1–S18), and X-ray diffraction data (Tables S1, S2 and Fig. S19–S24). CCDC 1560854 and 1560855. For ESI and crystallographic data in CIF or other electronic format see DOI: 10.1039/c7dt03010j A.E. Arbuzov Institute of Organic and Physical Chemistry, Kazan Scientific Center, Russian Academy of Sciences, Kazan, Russia. E-mail: elli@iopc.ru


NMR-experiments 4
Figure S1. 1D 1 H and 1 H{ 31 P} NMR spectra of 2 in C 6 D 6 at T = 303 K. 4 Figure S2. 1D 13 C{ 1 H}, 13 C DEPT, 31P and 31 P{ 1 H} NMR spectra of 2 in C 6 D 6 at T = 303 K. 5 Figure S3. 2D 1 H-1 H COSY NMR spectra of 2 in C 6 D 6 at T = 303 K. 6 Figure S4. 2D 1 H-13 C HSQC NMR spectra of 2 in C 6 D 6 at T = 303 K. 7 Figure S5. 2D 1 H-13 C HMBC NMR spectra of 2 in C 6 D 6 at T = 303 K. 8 Figure S6. 2D 1 H-31 P HMBC NMR spectra of 2 in C 6 D 6 at T = 303 K. 9  Table S1. Crystallographic data for 2 and 3 24 Table S2. Selected geometrical parameters of molecules 2 and 3 26 All manipulations were carried out with standard high-vacuum and dry-nitrogen techniques. Solvents were dried and degased prior to use and stored under nitrogen atmosphere. The ESI mass spectra were obtained on a Bruker Esquire 3000 Plus. The melting points were determined on a Boetius apparatus and are uncorrected.
Synthesis of starting 1,2-bis(phenylphosphino)ethane was carried by described methods: All other reagents were purchased from commercial sources and used as received 1 .

NMR-experiments.
All NMR experiments were performed with a Bruker AVANCE-500 spectrometer equipped with a 5 mm diameter gradient inverse broad band probehead and a pulsed gradient unit capable of producing magnetic field pulse gradients in the z-direction of 53.5 G·cm −1 . Frequencies are 500.13 MHz in 1 H NMR, 202.5 MHz in 31 P NMR and 125.8 MHz in 13 C NMR experiments. For 1 H-31 P long range correlations HMBC 2-3 experiment optimized for J = 10 Hz. NOE experiments were performed with 1D DPFGNOE 4 techniques. 1 H DOSY experiments were performed with ledbpgp2s 5 , using stimulated echo sequence and two spoil gradients. 31 P DOSY experiments were performed with stebpgpin1s, using stimulated echo sequence, INEPT for non-selective polarization transfer and one spoil gradient with decoupling during acquisition. Chemical shifts are reported in the δ (ppm) scale relative to the 1 H (7.15 ppm) and 13 C (128.6 ppm) signals of C 6 D 6 and 1 H (7.24 ppm) and 13 C (77.2 ppm) signals of CDCl 3 . 31 P chemical shifts were referenced to the 31 P signal of 85% H 3 PO 4 (0.00 ppm).

X-Ray diffraction.
A single crystals suitable for X-ray analysis were grown from diethylether 2 and chloroform 3 solutions.
Data of 2 and 3•were collected on a Bruker Smart Apex II CCD diffractometer using graphite monochromated MoK (λ = 0.71073 Å) radiation and -scan rotation. Data collection images were indexed, integrated, and scaled using the APEX2 data reduction package 6 and corrected for absorption using SADABS 7 . The structure was solved by direct methods and refined using SHELX program 8 . All nonhydrogen atoms were refined anisotropically. H atoms were calculated on idealized positions and refined as riding atoms. The inter-and intramolecular short contacts were evealuated by PLATON 9 . Crystal data and experimental details are given in Table S1. CCDC 1560854-1560855 (compounds 2 and 3 correspondingly) contain the supplementary crystallographic data for this paper. These data can be obtained free of charge via www.ccdc.cam.ac.uk/conts/retrieving.html (or from the Cambridge Crystallographic Data Centre, 12 Union Road, Cambridge CB2 1EZ, UK; fax: (+44) 1223-336-033; or deposit@ccdc.cam.uk).