Low-Coordinate Erbium(III) Single-Molecule Magnets with Photochromic Behavior

The structures and magnetic properties of photoresponsive magnets can be controlled or fine-tuned by visible light irradiation, which makes them appealing as candidates for ternary memory devices: photochromic and photomagnetic at the same time. One of the strategies for photoresponsive magnetic systems is the use of photochromic/photoswitchable molecules coordinated to paramagnetic metal centers to indirectly influence their magnetic properties. Herein, we present two erbium(III)-based coordination systems: a trinuclear molecule {[ErIII(BHT)3]3(dtepy)2}.4C5H12 (1) and a 1D coordination chain {[ErIII(BHT)3(azopy)}n·2C5H12 (2), where the bridging photochromic ligands belong to the class of diarylethenes: 1,2-bis((2-methyl-5-pyridyl)thie-3-yl)perfluorocyclopentene (dtepy) and 4,4′-azopyridine (azopy), respectively (BHT = 2,6-di-tert-butyl-4-methylphenolate). Both compounds show slow dynamics of magnetization, typical for single-molecule magnets (SMMs) as revealed by alternating current (AC) magnetic susceptibility measurements. The trinuclear compound 1 also shows an immediate color change from yellow to dark blue in response to near-UV irradiation. Such behavior is typical for the photoisomerization of the open form of the ligand into its closed form. The color change can be reversed by exposing the closed form to visible light. The chain-like compound 2, on the other hand, does not show significant signs of the expected trans–cis photoisomerization of the azopyridine in response to UV irradiation and does not appear to show photoswitching behavior.

S5 Figure S3. In-phase (χ') and out-of-phase (χ'') AC susceptibilities for 1 at 2.5 K measured in the 0-5000 Oe magnetic field range (a) and at 800 Oe in the 2.2-6.4 K temperature range (b). Values of α and τ parameters are gathered in Table S2. The solid lines are the best fits to generalized Debye model. Table S2. Values of α and τ parameters obtained from generalized Debye model fitting of the τ(ν) dependences for 1 at 2.5 K measured in the 0-5000 Oe magnetic field range depicted in Figure S3a (columns 1-5) and at 800 Oe in the 2.2-6.4 K temperature range depicted in Figure S3b (columns 6-10). T = 2.5 K ( Figure S3a) H = 800 Oe ( Figure S3b)  Figure 10a and 10c in the main text, respectively.  Figure S4. In-phase (χ') and out-of-phase (χ'') AC susceptibilities for 1UV at 2.5 K measured in the 0-5000 Oe magnetic field range (a) and at 1000 Oe in the 1.8-6.4 K temperature range (b). Values of α and τ parameters are gathered in Table S3. The solid lines are the best fits to generalized Debye model. Table S4. Values of α and τ parameters obtained from generalized Debye model fitting of the τ(ν) dependences for 1UV at 2.5 K measured in the 0-5000 Oe magnetic field range depicted in Figure S4a (columns 1-5) and at 1000 Oe in the 1.8-6.4 K temperature range depicted in Figure S4b (columns 6-10).   Figure 10a and 10c in the main text, respectively.
Magnetic field dependence (blue line in Figure 10a) Figure 10c) Figure S5. In-phase (χ') and out-of-phase (χ'') AC susceptibilities for 1vis at 2.5 K measured in the 0-5000 Oe magnetic field range (a) and at 1000 Oe in the 1.8-6.4 K temperature range (b). Values of α and τ parameters are gathered in Table S6. The solid lines are the best fits to generalized Debye model. Table S6. Values of α and τ parameters obtained from generalized Debye model fitting of the τ(ν) dependences for 1vis at 2.5 K measured in the 0-5000 Oe magnetic field range depicted in Figure S5a (columns 1-5) and at 1000 Oe in the 1.8-6.4 K temperature range depicted in Figure S5b (columns 6-10).  Table S8 (a) and S9 (b and c). The solid lines are the best fits to generalized Debye model. S12 Table S8. Values of α and τ parameters obtained from generalized Debye model fitting of the τ(ν) dependences for 2 at 1.8 K measured in the 100-5000 Oe magnetic field range depicted in Figure S6a.  Table S9. Values of α and τ parameters obtained from generalized Debye model fitting of the τ(ν) dependences for 2 at 500 Oe in the 1.8-3.4 K temperature range depicted in Figure S6b (columns 1-5) and at 2000 Oe in the 1.8-3.6 K temperature range depicted in Figure S6c (columns 6-10).   Figure 10b and 10d in the main text, respectively.
Magnetic field dependence (black lines in Figure 10b) branch 1 branch 2 T (K) 1 A (s -1 ) 0 (fixed) 733(65) B 2 (s -1 K -1 Oe -1 ) 0 (fixed) C (s -1 K -n ) 6(1) n 5.7(2) R 2 0.9794 Figure S7. In-phase (χ') and out-of-phase (χ'') AC susceptibilities for 2UV at 1.8 K measured in the 100-5000 Oe magnetic field range (a), at 500 Oe in the 1.8-3.4 K temperature range (b) and at 2000 Oe in the 1.8-3.6 K temperature range. Values of α and τ parameters are gathered in Table S11 (a) and S12 (b and c). The solid lines are the best fits to generalized Debye model. S15 Table S11. Values of α and τ parameters obtained from generalized Debye model fitting of the τ(ν) dependences for 2UV at 1.8 K measured in the 100-5000 Oe magnetic field range depicted in Figure S7a.  Table S12. Values of α and τ parameters obtained from generalized Debye model fitting of the τ(ν) dependences for 2UV at 500 Oe in the 1.8-3.4 K temperature range depicted in Figure S7b (columns 1-5) and at 2000 Oe in the 1.8-4.0 K temperature range depicted in Figure S7c (columns 6-10), respectively.   Figure 10b and 10d in the main text, respectively.
Magnetic field dependence (blue lines in Figure 10b

Preparation of (4-bromo-5-methylthiophen2-yl)boronic acid (C)
The following reaction was performed under inert gas atmosphere. 3,5-dibromo-2-methylthiophene (B) (9.7 g, 37.9 mmol, 1 eq) was dissolved in anhydrous THF in a two-necked round bottom flask. The mixture was cooled to -78 °C. To the stirred solution 2.5 M n-BuLi (16 ml, 39.8 mmol, 1.5 eq) was added dropwise. The reaction was stirred for 15 minutes at -78 °C and then tributyl borate (10.46 g, 45.5 mmol, 1.2 eq) was slowly added to the mixture followed by slow warm up to RT and further stirring at RT for 24h. The reaction mixture was quenched with 110 ml of 1M HCl and stirred for additional 1 h. Then THF was removed under vacuum and the mixture was extracted with Et 2 O (3 × 50 ml). The organic layers were combined and the boronic acid was removed by washing with 2M NaOH solution (3 × 50 ml). The combined aqueous layers were carefully acidified to neutral using conc. HCl. The obtained off-white-toyellow solid was filtered and dried under vacuum (yield: 5.5 g, 76%) and used without purification for the preparation of 4-(4-bromo-5-methylthiophen-2-yl)pyridine (D) as described in the next step described below. Crude 1 H NMR (300 MHz, DMSO; Figure S9) δ 8.28 (s, 2H), 7.50 (s, 1H), 2.37 (s, 3H). Figure S9. 1 H NMR spectrum of (4-bromo-5-methylthiophen2-yl)boronic acid (C) recorded in CDCl 3 with TMS as the reference.

Preparation of 4-(4-bromo-5-methylthiophen-2-yl)pyridine (D)
The (4-bromo-5-methylthiophen2-yl)boronic acid (C) (4.0 g, 18.1 mmol, 1 eq), 4-iodopyridine (4.45 g, 21.7 mmol, 1.2 eq) and 50 ml THF were placed in a 200 ml Schlenk flask udner inert atmosphere. Into the stirred solution under argon 50 ml of 20% Na 2 CO 2 solution in degassed water was added and the reaction mixture was degassed again by performing three vacuum-inert gas cycles. Pd(PPh 3 ) 4 (1.04 g, 0.9 mmol, 5 mol%) was added and the reaction mixture was heated to 85 °C for 72 h under inert gas. After cooling to RT, the product was extracted from the mixture by two 50 ml portions of dichloromethane and the organic layers were washed with H 2 O (2 × 50 ml) and brine (50 ml). After drying over MgSO 4 the solvent was removed under vacuum yielding an orange solid as the crude product. Purification by silica gel column chromatography (EtOAc) afforded a light yellow solid (4.32 g, 93% yield) as the final product. 1 H NMR (300 MHz, CDCl 3 ; Figure S10