Engineering the rhenium metalation of phenanthroline imino complexes for variable structure activity relationships

Abstract

There are marked advantages for incorporating multiple metal atoms into a single entity specifically for theranostic and radiopharmaceutical development for the treatment and imaging of cancer. To this end, we have synthetically manipulated a series of rhenium tricarbonyl complexes based on a 1,10-phenanthroline imino bifunctional ligand system to allow for the formation of mononuclear, dinuclear and mixed di-metal rhenium tricarbonyl complexes. Thus, diverse protein bio-chemical interactions are in principle possible by harnessing the chemical versatility of the the phenanthroline scaffold together with strategic metal substitution to advance drug development. The structure–activity relationships for these complexes are described indicating covalent and non-covalent interactions which are explored through the lens of macromolecular interactions and metallo-inorganic drug design. Complex characterisation was conducted with IR, ¹H and ¹³C NMR, EA and detailed structural investigations using SC-XRD. Hirshfeld surface analysis indicated that halogen, hydrogen and oxygen-based hydrogen interactions play dominant roles in crystal packing and are significant to consider for drug design. In silico drug likeness analysis of the compounds using Swiss ADME had predicted good potential bioactivity. HeLa and MCF-7 cell SRB assay was used to identify the cytotoxic properties of the fac-tricarbonyl-(1,10-phenanthroline-N,N′)-(bromido)-rhenium(I), fac-tricarbonyl-{2-([1,10-phenanthrolin-5-ylimino]methyl)phenol-N,N′}-(bromido)-rhenium(I), fac-tricarbonyl-{2-([1,10-phenanthrolin-5-ylimino]methyl)phenol-N,N′}-(chlorido)-rhenium(I), fac-{2-([1,10-phenanthrolin-5-ylimino]methyl)phenolato-N,N′}-(methanol)(bromido)-hexacarbonyl-di-rhenium(I), and fac-bis{2-([1,10-phenanthrolin-5-ylimino]methyl)phenol-N,N′}-hexacarbonyl-di-rhenium(I). Excellent activity was shown for the free ligand and 1,10-phenanthroline with IC₅₀ values of 9 ± 3 μM, 11 ± 2 μM (HeLa) and 8 ± 2 μM, 6 ± 2 μM (MCF-7) respectively. Rhenium complexes showed IC₅₀ values ranging between 15 ± 2 μM and 35 ± 2 μM against HeLa cells and between 68 ± 2 μM and >100 μM against MCF-7 cells. A correlation is observed between good to moderate cytotoxicity and a balanced spread of H-bond interactions showing amphiphilic character.

---

1 Introduction

The treatment of cancer has proven for a sustained period of time to be a challenging subject, where even in ancient times malignant tumours were known but considered untreatable.¹ Fortunately at present, multiple options are available to treat or image neoplastic tissue. Treatment protocols can vary and may involve the surgical removal of a solid tumour mass, chemical anti-tumour drugs and/or radiological methods such as external beam radiation.² Chemotherapeutic agents, including radiopharmaceuticals have historically proved valuable and remain the frontline choice for advanced-stage malignancies where surgery and/or external radiation therapy cannot be prescribed due to various reasons.³ Treatment regimens can leverage the cytotoxic properties of a compound, prime examples being cisplatin and doxorubicin, which have different interaction modes with DNA yet inhibit the macromolecular processes resulting in cancer cell death (Fig. 1).^4,5

Fig. 1 Illustrative manners which cisplatin and doxorubicin, the well-known anti-cancer compounds, may bind with DNA. To the left is cisplatin covalently bound to DNA (PDB code: 1AIO). The line drawing of cisplatin indicated in the inset-coloured green. To the right is the organic compound, doxorubicin interacting non-covalently with DNA (PDB code: 151D). Chemical formula of Doxorubicin illustrated in blue.

Two modes of small molecule-to-protein interactions, as exemplified by cisplatin and doxorubicin, are direct covalent versus non-covalent binding and is a key aspect to consider for metallo-protein binding. Several rhenium-containing protein structures show both covalent protein-metal binding and non-covalent π–π interactions between the metal-complex and the protein.⁶ For example fac-[Re(CO)₃(phen)]⁺ (phen = 1,10-phenanthroline) complex binds covalently with azurin at HIS83 (PDB Code: 1JZI)⁷ and fac-[Re(CO)₃(dmp)]⁺ (dmp = 4,7-dimethyl-1,10-phenanthroline) bind covalently at HIS124 (PDB Code: 2I7O)⁸ with additional π–π interactions occurring between the metal complex and the TRP122 residue of the parent protein. Further π–π interactions are observed between the metal complex and its symmetry related counterpart. These non-covalent and covalent interactions are illustrated in Fig. 2. In many instances, defining the small molecule to protein interactions (both covalently and weak) in a macromolecular environment is considered valuable for the design future lead-like compounds. The inverse may similarly be true, that examining the dominant interactions of small molecules in their respective crystal structures could well indicate key interactions which may occur between the small molecule and protein.

Fig. 2 (A) An illustration of the covalent binding of fac-[Re(CO)₃(phen)]⁺ with azurin at the HIS83 site. (B) The non-covalent interactions of fac-[Re(CO)₃(dmp)]⁺ and the side chain of TRP122 as well as the π–π interaction of the metal complex with its symmetry(*) counterpart.

Our interest is in the development of metallo-drugs for cancer treatment and understanding the metal protein binding that would occur. Complexes, containing one or more metals are well suited for theranostic application, which was defined by Duan et al. (2022) in the statement: ‘See what you treat and treat what you see…’.⁹ The rational being that a medical procedure can both treat and diagnose a particular ailment at the same time, in real time. Radio-theranostics utilises radionuclides to emit both therapeutic and diagnostic radiation,¹⁰ thus combining two radionuclides into one compound.¹¹ The combination of the two radiation types has some desirable advantages over a step wise treatment regime. Notably, when diagnostic testing in conjunction with therapy occurs, the treatment site and cellular response to treatment can be measured immediately before marked anatomical changes takes place, thus identifying if target coordination to a disease site has occurred which tailors the treatment for patient's optimal recovery.¹²

Technetium and rhenium are archetypal for this purpose due to their Group 7 periodic nature. Technetium-99 m has ideal radioactive properties for diagnosis. The half-life is 6 hours with a gamma emission energy of 140 keV.^{13 99m}Tc complexes are currently utilised in the medical imaging of select organs such as: ^99mTc^VIIO₄⁻ for thyroid imaging, ^99mTc^V-tetrofosmin and ^99mTc^I-sestamibi for myocardial perfusion imaging, and renal imaging can be conducted using ^99mTc^V-Mercaptoacetyltriglycine.^{14 99m}Tc has the benefit of easy accessibility with the use of the ⁹⁹Mo/^99mTc generator making it suitable for use in clinical environments.¹⁵

Rhenium exists both in its ‘cold’ nonradioactive form, making it ideal for laboratory research as well as having two radionuclides appropriate for clinical therapy in the field of oncology. ¹⁸⁶Re and ¹⁸⁸Re have half-lives of 3.7 days for ¹⁸⁶Re and 17 hours for ¹⁸⁸Re¹⁶ and emit β_max⁻ emission energy of 1070 keV and a γ-ray emission energy of 137 keV for ¹⁸⁶Re¹⁷ while ¹⁸⁸Re has a β_max⁻ emission energy of 2120 keV and a γ-ray emission energy of 155 keV.¹⁸ Several rhenium compounds have found use in specific clinical applications. ¹⁸⁶Re-sulphide-colloid is effective for the treatment of rheumatoid arthritis¹⁹ and ¹⁸⁶Re-HEDP for bone metastases,²⁰ while select examples for ¹⁸⁸Re complexes are ¹⁸⁸Re-HDD-iodized oil, ¹⁸⁸Re-HDD-lipiodol and ¹⁸⁸Re-HSA microspheres which have been employed for hepatocellular carcinoma,^21–23 and skin cancer can be treated with ¹⁸⁸Re-SCT.²⁴

Structure activity relationship (SAR) studies attempt to correlate the resultant biological activity to the chemical and structural characteristics of a compound²⁵ and thus each portion of the metal complexes investigated in this study have characteristics advantageous for metallodrug development. The rhenium(I) tricarbonyl core is considered stable in biological environments (and suitable for biological applications²⁶) due to the low d⁶ spin of the rhenium in the +1-oxidation state.²⁷ Oxygens found on the periphery of transition metal complexes (due to the presence of the M–C≡O carbonyl groups) gives access to hydrogen bonding donor atoms²⁸ which may promote interaction with the hydrophilic moieties in a biological environment, such as solvent channels or hydrophilic residues within a protein. Bidentate ligand systems can facilitate non-covalent complex-to-protein interactions (via the organic ligand) versus the final octahedral open site of the rhenium metal which can then accommodate direct covalent metal–protein coordination.²⁹

We have therefore engineered a series of multinuclear rhenium complexes taking inspiration from the wwPDB which indicated metal–protein covalent bonding; non-covalent ligand π–π interactions and SAR relationships and merged it with the potential of theranostic application containing multiple metal centres both in a symmetrical (i.e. dimer) and asymmetrical manner. To achieve a stepwise synthetic approach, our preference relied on the use of imino-based functionalised bidentate ligand systems, obtained via the Schiff-base reaction. These are prevalent in nature and are present in many biological organisms.³⁰ Two prominent examples are transamination and retinaldehyde binding in the retinoid cycle.³¹ The imine bond is a useful pharmacophore easily adaptable to leverage a particular biological activity,³² and have been proposed for use in anti-cancer studies.³³ The aldehyde portion of the Schiff-base used in the study, is a methylated form of salicylic aldehyde which in turn is related to salicylic acid, a major metabolite of aspirin.³⁴ Aspirin itself is useful for its analgesic properties making salicylic aldehydes an interesting group for study. The amine counterpart (1,10-phenanthroline-5-amine) has strong metal binding affinity³⁵ and the 1,10-phenanthroline functional group can have non-covalent (i.e. π–π) interactions in a biological setting (Fig. 2) potentially tuning the metal complex for various protein interactions.³⁶ Our complexes have an open metal coordination site available (represented by the Re–Br bond, Fig. 3) and thus Re-halogen substitution can readily occur, which in protein environment would subsequently allow for metal–protein covalent interactions.^37,38 A graphical description of the potential SAR which motivated the design of this study is illustrated in Fig. 3.

Fig. 3 Schematic representation of the engineered complexes, with the potential SAR of the metal complex illustrated for representative portions of the complex.

Furthermore, our compounds have the potential for future theranostic applications (technetium + rhenium) through the coordination of a second metal to the salicylidene group on the ligand. This may occur in a symmetrical manner, to form the dinuclear complexes achieved via dimerization,³⁹ or in an asymmetrical manner if the second metal coordinates to the N,O portion of the ligand.

We present here, the synthetic manner which resulted in this series of rhenium tricarbonyl 1,10-phenantroline based complexes which incorporated the SAR relationships and potential theranostic application described above. Four complexes utilize the modified 5-methyl-2-{(E)-[(1,10-phenanthrolin-5-yl)imino]methyl}phenol (L) bidentate ligand^40,41 to coordinate to the metal and compared relative to the standard 1,10-phenanthroline. The mononuclear rhenium complexes (1–3), the symmetrical dinuclear specie formed from dimerization (5) and asymmetrical di-metal complex with metal atoms on both sides of the ligand (4) were analysed by infrared, ¹H and ¹³C nuclear magnetic resonance spectroscopy, and elemental analysis. Single crystal X-ray diffraction experiments illustrated the coordination geometry of each complex and Hirshfeld surface analysis was used to evaluate the dominant small molecule interactions present with consideration of the interactions possible between small and macromolecular structures. The metal complexes were tested against HeLa and MCF-7 cells to determine their cytotoxic potential. Line drawing of the complexes investigated in this study are illustrated in Fig. 4.

Fig. 4 Line drawings of the complexes investigated in this study namely: fac-tricarbonyl-(1,10-phenanthroline-N,N′)-(bromido)-rhenium(I) (1), fac-tricarbonyl-{2-([1,10-phenanthrolin-5-ylimino]methyl)phenol-N,N′}-(bromido)-rhenium(I) (2), fac-tricarbonyl-{2-([1,10-phenanthrolin-5-ylimino]methyl)phenol-N,N′}-(chlorido)-rhenium(I) (3), fac-{2-([1,10-phenanthrolin-5-ylimino]methyl)phenolato-N,N′}-(methanol)(bromido)-hexacarbonyl-di-rhenium(I) (4), and fac-bis{2-([1,10-phenanthrolin-5-ylimino]methyl)phenol-N,N′}-hexacarbonyl-di-rhenium(I)] (5).

2 Experimental

2.1 Materials, synthesis, and crystallization

All reagents and solvents used for this study were of analytical grade, purchased from Sigma-Aldrich and utilized without further purification. Rhenium pentacarbonyl bromide [Re(CO)₅Br] was purchased from Strem Chemicals, Newburyport, Massachusetts, USA. Rhenium pentacarbonyl chloride [Re(CO)₅Cl] was purchased from Sigma-Aldrich and used as is. fac-[NEt₄]₂[Re(CO)₃Br₃] (ReAA) was synthesised according to the procedure found in literature.⁴² The ligand 5-methyl-2-{(E)-[(1,10-phenanthrolin-5-yl)imino]methyl}phenol (L) 5Me-SalH-Phen was synthesized according to literature with an adjusted protocol (methanolic solution under reflux conditions for three hours).⁴¹

The ¹H and ¹³C NMR spectra (150.96 and 300.13 MHz respectively) of the ligands and complexes were recorded on a Bruker 400 MHz AVANCE III NMR spectrometer at 25 °C in (CD₃)₂SO (2.50 ppm). Positive shifts are downfield. All chemical shifts (δ) are reported in ppm and the coupling constants in Hz. A Bruker Tensor 27 Standard System spectrophotometer equipped with a 4000–370 cm⁻¹ laser range was used to record the infrared spectra. All data were collected at room temperature. Elemental analysis was conducted at the University of KwaZulu Natal, South Africa (238 Mazisi Kunene Road, Glenwood, Durban, 4041) and Atlantic Microlab, Inc. North America (6180 Atlantic Blvd. Suite M Norcross, GA 30071).

2.1.1 fac-Tricarbonyl-(1,10-phenanthroline-N,N′)-(bromido)-rhenium(I) – fac-[Re(CO)₃(Phen)(Br)] (1). Rhenium pentacarbonyl bromide (0.0994 g, 0.2447 mmol) was dissolved in methanol (10 ml). The 1,10-phenanthroline (0.0444 g, 0.2464 mmol) was dissolved in methanol (10 ml) and added to the metal solution. The mixture was allowed to reflux overnight at 90 °C. The product precipitated as a bright yellow powder (yield: 0.0912 g, 70%). Crystals suitable for single crystal diffraction were grown from deuterated DMSO. Elemental analysis for C₁₅H₈BrN₂O₃Re (1) found (calculated) (%): C 34.03 (33.97), H 1.40 (1.52), N 5.16 (5.28). IR (ATR cm⁻¹): ν(C≡O) 2014, 1926, 1887. ¹H NMR (400 MHz, DMSO-d₆): δ 9.46 (q, 2H J = 5.1, 1.3 Hz, Ar), 8.97 (q, 2H J = 8.2, 1.3 Hz, Ar), 8.34 (s, 2H, Ar), 8.11 (q, 2H J = 8.2, 5.1 Hz, Ar). ¹³C NMR (400 MHz, DMSO-d₆): δ 153.7, 145.8, 139.3, 130.6, 127.8, 126.6 (Ar).

2.1.2 fac-Tricarbonyl-{2-([1,10-phenanthrolin-5-ylimino]methyl)phenol-N,N′}-(bromido)-rhenium(I) – fac-[Re(CO)₃(Phen5MeSalH)(Br)] (2). Rhenium pentacarbonyl bromide (0.0968 g, 0.2383 mmol) and (L) 5Me-SalH-Phen (0.0746 g, 0.2381 mmol) were dissolved in benzene (10 ml). The mixture was allowed to reflux overnight at 80 °C. The product precipitated out as a yellow powder (yield: 0.1282 g, 81%). Elemental analysis for C₂₃H₁₅BrN₃O₄Re (2) (with 4.9% [Re(CO)₅Br] unreacted starting complex) found (calculated) (%): C 40.59 (40.33), H 2.11 (2.17), N 6.10 (6.03). IR (ATR cm⁻¹): ν(C≡O) 2022, 1908, 1874; ν(C=N) 1610, 1595. ¹H NMR (400 MHz, DMSO-d₆): δ 11.78 (s, 1H, OH), 9.51 (d, 1H J = 4.7 Hz, Ar), 9.38 (d, 1H J = 4.7 Hz, Ar), 9.14 (s, 1H, Ar), 9.04 (d, 1H J = 8.4 Hz, Ar), 8.89 (d, 1H J = 8.2 Hz, Ar), 8.16–8.06 (m, 3H, Ar), 7.81 (d, 1H J = 8.0 Hz, Ar), 7.36 (s, 1H, Ar), 6.89 (s, 1H, Ar), 6.88 (s, 1H, Ar), 2.37 (s, 3H, Ar). ¹³C NMR (400 MHz, DMSO-d₆): δ 164.9, 160.1, 154.1, 152.6, 146.9, 146.2, 145.5, 144.6, 138.9, 135.4, 131.5, 130.7, 127.5, 126.8, 126.7, 120.9, 117.8, 117.0, 113.3 (Ar), 21.5 (CH₃).

2.1.3 fac-Tricarbonyl-{2-([1,10-phenanthrolin-5-ylimino]methyl)phenol-N,N′}-(chlorido)-rhenium(I) – fac-[Re(CO)₃(Phen5MeSalH)(Cl)] (3). Rhenium pentacarbonyl chloride (0.1023 g, 0.2828 mmol) was dissolved in benzene (10 ml). To the metal solution, a solution of benzene (10 ml) and (L) 5Me-SalH-Phen (0.0885 g, 0.2824 mmol) was added, and the mixture allowed to reflux overnight at 90 °C. The product precipitated out as a yellow precipitate (yield: 0.1461 g, 84%). Small yellow crystals, suitable for SCXRD, were obtained by slow evaporation of a benzene solution. Elemental analysis for C₂₃H₁₅ClN₃O₄Re (3) (with 4.6% [Re(CO)₅Cl] unreacted starting complex) found (calculated) (%): C 42.73 (43.32), H 2.23 (2.33), N 6.35 (6.47). IR (ATR cm⁻¹): ν(C≡O) 2026, 1908, 1878, 1860; ν(C=N) 1609, 1596. ¹H NMR (400 MHz, DMSO-d₆): δ 11.77 (s, 1H, Ar), 9.49 (d, 1H J = 4.8 Hz, Ar), 9.35 (d, 1H J = 4.8 Hz, Ar), 9.13 (s, 1H, Ar), 9.04 (d, 1H J = 8.4 Hz, Ar), 8.89 (d, 1H J = 8.2 Hz, Ar), 8.16–8.05 (m, 3H, Ar), 7.80 (d, 1H J = 7.8 Hz, Ar), 6.89 (s, 1H, Ar), 6.87 (s, 1H, Ar), 2.36 (s, 3H, CH₃). ¹³C NMR (400 MHz, DMSO-d₆): δ 164.9, 160.0, 153.9, 152.4, 146.8, 146.2, 145.5, 144.6, 139.0, 135.4, 131.5, 130.6, 127.5, 126.8, 126.7, 120.9, 117.8, 117.0, 113.2 (Ar), 21.5 (CH₃).

2.1.4 fac-{2-([1,10-Phenanthrolin-5-ylimino]methyl)phenolato-N,N′}-(methanol)(bromido)-hexacarbonyl-di-rhenium(I) – fac-[Re(CO)₃Br(Phen5MeSal)Re(CO)₃(MeOH)] (4). Rhenium pentacarbonyl bromide (0.1503 g, 0.3701 mmol, 2 : 1 – Re : L equivalents) was dissolved in methanol (15 ml). The ligand (L) 5Me-SalH-Phen (0.0580 g, 0.1851 mmol) was also dissolved in methanol (15 ml) and NaOH (0.0078 g, 0.1950 mmol) was added to deprotonate the ligand. The ligand and metal solutions were combined and allowed to reflux for 2 days. The product formed as a yellow precipitate (yield: 0.1253 g, 70%). Small crystals were grown from methanol which were suitable for single crystal X-ray diffraction. The product dissolves sparingly in most solvents. Elemental analysis for C₂₇H₁₈BrN₃O₈Re₂ (3) (with 4.2% [Re(CO)₅Br] unreacted starting complex) found (calculated) (%): C 32.65 (32.85), H 1.66 (1.80), N 4.89 (4.18). IR (ATR cm⁻¹): ν(C≡O) 2018, 1920, 1892; ν(C=N) 1641, 1619. ¹H NMR (400 MHz, DMSO-d₆): δ 9.56 (s, 1H, Ar), 9.45 (s, 1H, Ar), 9.07–8.65 (m, 3H, Ar), 8.21–7.97 (m, 3H, Ar), 7.28 (dt, 1H J = 37.6, 7.4 Hz, Ar), 6.80 (s, 1H, HC=N), 6.53 (t, 1H J = 6.8 Hz, Ar), 2.28 (s, 3H, CH₃). Solution state NMR indicates the presence (<5%) of the ligand dissociated fac-[Re(CO)₃(PhenNH₂)Br] complex. ¹³C NMR (400 MHz, DMSO-d₆): δ 197.1, 168.8, 154.5, 153.7, 153.5, 151.8, 147.6, 145.9, 144.3, 139.3, 136.5, 129.6, 127.2, 127.0, 126.6, 122.0, 119.2, 117.9, 117.1 (Ar), 21.7 (CH₃).

2.1.5 fac-Bis{2-([1,10-phenanthrolin-5-ylimino]methyl)phenol-N,N′}-hexacarbonyl-di-rhenium(I) – fac-[Re(CO)₃(Phen5MeSal)]₂ (5). (L) 5Me-SalH-Phen (0.0863 g, 0.2754 mmol) was dissolved in methanol (10 ml). To the ligand solution, NaOH (0.0112 g, 0.2801 mmol) was added, and the mixture heated and stirred for 10 minutes. Rhenium pentacarbonyl chloride (0.1001 g 0.2767 mmol, 1 : 1 – Re : L equivalents) was dissolved in methanol (10 ml) and added to the basic ligand solution. The mixture was allowed to reflux over two days. The product was filtered off as an orange precipitate (yield: 0.1233 g, 76%). Crystals suitable for single crystal X-ray diffraction were grown from DMF. The product dissolves only sparingly in most solvents. Elemental analysis for C₄₆H₂₈N₆O₈Re₂ (5) (with 4.2% [Re(CO)₅Cl] unreacted starting complex) found (calculated) (%): C 46.41 (46.11), H 2.30 (2.32), N 6.92 (6.91). IR (ATR cm⁻¹): ν(C≡O) 2013, 1888, 1875; ν(C=N) 1608, 1578. ¹H NMR (400 MHz, DMSO-d₆): δ 9.51 (d, 1H J = 4.9 Hz, Ar), 9.37 (d, 1H J = 4.8 Hz, Ar), 9.15 (s, 1H, Ar), 9.05 (d, 1H J = 8.4 Hz, Ar), 8.89 (d, 1H J = 8.2 Hz, Ar), 8.15 (q, 1H J = 4.4 Hz, Ar), 8.11 (s, 1H, Ar), 8.07 (q, 1H J = 4.4 Hz, Ar), 7.82 (d, 1H J = 8.0 Hz, Ar), 6.89 (s, 1H, HC=N), 6.88 (d, 1H J = 7.1 Hz, Ar), 2.37 (s, 3H, CH₃).

2.2 X-ray structure determination and refinement

The reflection data were collected on Bruker D8 Venture 4K Kappa Photon III C28 diffractometer and Bruker D8 Quest diffractometers. Both diffractometers are equipped with a graphite monochromator using a Mo-Kα X-ray generator with a wavelength of λ = 0.71073 Å. On the Bruker D8 Venture, a second X-ray generator is equipped with a Cu-Kα anode producing X-ray at λ = 1.54178 Å. Data was collected utilizing both phi and omega scans at a temperature of 100 K. COSMO⁴³ was utilized for multiple hemisphere data collection of the reciprocal space. Bruker SAINT-Plus and XPREP⁴⁴ were employed for frame integration and data reduction respectively. SADABS⁴⁵ was used for intensity and absorption correction through the multi-scan method. SHELXT⁴⁶ was used to solve the crystal structures through intrinsic phasing. WinGX,⁴⁷ Olex2⁴⁸ and SHELXL-2018/3⁴⁶ was used for the refinement of the crystal structures. The DIAMOND 4.0⁴⁹ software was utilized for the generation of images. Thermal ellipsoids are drawn with 50% probability level if not stated otherwise. In all structures the hydrogen atoms were positioned geometrically and refined using a riding model: C–H aromatic distances at 0.95 Å; methine C–H distances at 1.00 Å; methylene C–H distances at 0.99 Å and methyl C–H distances at 0.98 Å. The H atom isotropic displacement parameters were fixed at U_iso(H) = 1.2 U_eq(C). The hydrogen atoms bound to non-carbon atoms were placed according to the Fourier electron density difference map. In each image the hydrogen labels have been removed for the sake of clarity unless used to indicate specific features. When applicable each structure's inter- and intramolecular hydrogen bonding are indicated in green dashed lines. A low occupancy carbonyl – halide disorder is possibly present (<5%) in (1) and were evaluated by several refinement strategies. Free refining the atoms or refining the disorder yielded NPD displacement parameters as well as unstable atom positions, even with the use of constraints, leading to fragmentation of the molecule and high shift values. The non-disordered structure was selected as the best and most stable representative of the crystal data, supported by the CheckCIF guidelines. Complex (4) was found to be modulated with two domains that could be defined (being −178.88° rotated with respect to each other). The first domain has a total of 44 175 reflections and 4313 unique reflections where the second domain has 42 329 reflections and 4307 unique reflections, these values were obtained before final unit cell determination. Only the first domain was used for structure solution and refinement. Representative photographs of the external crystal morphology of (1), (3), (4), and (5) as viewed through a polarizing microscope with 45× magnification can be found in the SI.

2.3 Cytotoxicity testing

Cellonex human cervical cancer cell line (HeLa cells), as well as estrogen- and progesterone-positive luminal breast cancer cell line (MCF-7) was grown at 37 °C in a humidified atmosphere containing 5% carbon dioxide. Dulbecco's modified essential medium supplemented with 10% fetal calf serum and 1% penicillin/streptomycin was used as growth medium for both cell lines. Cytotoxicity tests were performed according to the SRB (Sulforhodamine B) assay.⁵⁰ Cells were trypsinized and suspended in 1 ml of growth medium. Cells were diluted to a cell count of 0.5 × 10⁵ cells per ml. Of this solution, 0.1 ml was added to each well of a 96 well microplate and incubated for to allow cells to adhere. HeLa cells were incubated for 1 hour, while MCF-7 cells were incubated overnight. Varied concentrations of the tested compounds (0.1 ml) were added to the cells. Plates were incubated again for 3 days at 37 °C with 5% CO₂. Cells were fixated by the addition of trichloroacetic acid (0.05 ml of 50%) and stored overnight at 4 °C. The plates were washed under running tap water and dried at 50 °C for 2 hours. SRB stain (0.1 ml) was added to each well and stored in the dark for 30 minutes. The unbound stain was removed by washing with 1% acetic acid (0.1 ml × 4) followed by air-drying overnight. Tris buffer (0.1 ml) was added to each well in order to solubilize the dye, followed by gentle shaking for 1 hour. Absorbance was measured at 510 nm. Growth inhibition of test compounds was determined from absorbance measurements as a percentage of the control. One-way ANOVA with Dunnett's post-test was performed using GraphPad Prism version 5.00 for Windows, GraphPad Software, San Diego California USA. Data was fitted to a nonlinear regression vs. normalized response. All experiments were carried out in triplicate (n = 3) and the results are presented as mean ± SEM (standard error of mean) values, listed in the SI.

2.4 Hirshfeld analysis

Hirshfeld analysis was conducted to quantitively determine the intermolecular interactions experienced by the metal complexes. These interactions can be represented from the normalized contact distance (d_norm) values which is calculated by Crystal Explorer 21.5,^51,52 and is calculated from the following equation:
where d_i is the distance from the nearest internal atom (I) to and on the inside of the Hirshfeld surface and d_e is the distance from the closest external atom (E) from and external to the Hirshfeld surface, r^vdW is the van der Waals radius of the atom in question. The d_norm values can be represented through plotting on a fingerprint plot and by colouring the Hirshfeld surface itself. The d_norm values can be plotted on the Hirshfeld surface and stronger interactions (such as hydrogen interactions) are indicated in red where weaker interactions (for example van der Waals interactions) are depicted in blue. The fingerprint plot graphs the d_i and d_e values on a two-dimensional plot to show the extent of type of interactions present in the structure. These interactions can be separated into atom type pairs to show the percentage contribution of each type of interactions with respect to the whole Hirshfeld surface. Since the Hirshfeld surface is calculated from the geometry of the crystal structures no quantum mechanical wavefunction calculations are required.⁵³ For each structure any disorder is removed and refined before surface generation is attempted. In each case the major disorder is used for the refinement.

3 Results and discussion

3.1 Synthesis and characterisation

The synthetic procedure of the rhenium complexes involves dissolving the ligand (L) in an appropriate solvent, normally methanol. The metal reagent, either [Re(CO)₅X] (where X = Cl or Br) is added to the ligand solution in a 1 : 1 (for complex 1–3, 5) or 2 : 1 (for 4) ratio. The mixture is then refluxed for 12–24 hours to ensure complete reaction with [Re(CO)₅X]. Generally, the [Re(CO)₅Cl] and [Re(CO)₅Br] reagents produce higher product yields and purity in comparison to when fac-[NEt₄]₂[Re(CO)₃Br₃] (ReAA) is used however longer reaction times are required. The advantage of these reagents is the lack of [NEt₄]Br salt which often presents challenges when attempting to separate it from the final product. The synthesis of the metal complexes with ReAA and (L) was attempted however this did not yield favourable results and so [Re(CO)₅X] were the precursors of choice. In the synthesis of compounds (2), (3), and (5) the metal to ligand (L) were reacted with an equivalent metal to ligand ratio (i.e. 1 : 1). Complex (4) has a similar synthetic procedure to (2) however the molar ratio of metal reagent to ligand is 2 : 1 resulting in the di-metal complex. The correct Re : L ratio and deprotonating the ligand for (4) and (5) were key synthetic aspects needed to control the di-metal versus dinuclear formation.

Use of the [Re(CO)₅Cl] precursor yielded crystals of (3), whereas (2) typically crashes out as a powder and alternative crystallisation attempts did not yield suitable crystals for analysis. Synthetic efforts were made to coordinate the rhenium exclusively to the N,O-salicylidene portion of the ligand but were unsuccessful, which was ascribed to the 1,10-phenantroline's high metal binding affinity leading to the preferential first coordination of the N,N′ donor atoms to the Re atom. The synthetic reaction scheme is illustrated in Fig. 5.

Fig. 5 Reaction scheme for the formation of the 5Me-SalH-Phen (L)-based, mononuclear (2, 3), di-metal (4) and dinuclear (5) rhenium complexes.

The key spectroscopic indicators from IR and ¹H NMR are listed Table 1 specifying the imine and carbonyl stretching frequencies, as well as C–OH ppm chemical shift values. From the IR spectroscopic analysis, the imine stretching frequencies for complex (2) and (3) has similar ν(C=N) frequencies of 1610, 1595 cm⁻¹ and 1609, 1596 cm⁻¹, which can be expected due to the identical ligand, and the variation of the halide (Br vs. Cl) on the sixth position of the metal. The presence of the halide is indicated by the shift in the carbonyl stretching frequency ν(C≡O) (2022 vs. 2026 cm⁻¹). Complex (4) has ν(C=N) frequencies of 1641 and 1619 cm⁻¹, and this can be attributed to the second metal binding to the salicylidene portion of the complex affecting the electron density on the imine bond. Complex (5) has similar ν(C=N) frequencies to (2) and (3) with values of 1608 and 1578 cm⁻¹ correlating well to the structure as no metal binding occurs on the imine bond. The non-halogenated complexes (4 & 5) have similar carbonyl stretching frequencies to (1).

Table 1 List of carbonyl and imine infrared (IR) stretching frequencies of complexes and ¹H NMR chemical shifts for complexes (1–5)

Compound	Stretching frequencies (cm^−1)		^1H NMR chemical shifts (ppm)
	ν(C=N)	ν(C≡O)	C–OH (measured in DMSO-d6)
(1)	—	2014, 1926, 1887	—
(2)	1610, 1595	2022, 1908, 1874	11.78
(3)	1609, 1596	2026, 1908, 1878	11.77
(4)	1641, 1619	2018, 1920, 1892	—
(5)	1608, 1578	2013, 1888, 1875	—

---

Nuclear magnetic resonance of the complexes, all conducted in DMSO-d₆, indicated the similarity of the spectra of (2) and (3) as expected due to their identical atomic coordination with the exception of the variation in the halogen atom. Also present is the free salicylidene C–OH peaks with 11.78 ppm and 11.77 ppm values respectively, which is not observed in complexes (4) and (5) due to metal coordination (SI Fig. S15). The spectra simplicity of (1) is due to the highly symmetrical nature of the structure.

3.2 Single crystal structure analysis

The single crystal structure analysis could be conducted for four of the rhenium 1,10-phenanthroline based complexes illustrating their absolute atomic configuration. For the mononuclear complexes (1) and (3), the metals are exclusively coordinated to the bidentate N,N′ 1,10-phenanthroline functional group/ligand, whereas (5) also indicates the formation of the dinuclear specie. Complex (4) has an additional Re metal bound to the N,O salicylidene functional group of 5Me-SalH-Phen (L) to form the asymmetrical di-metal specie. The coordination spheres of the metals are completed by a monodentate bromido ligand for (1) and (4), chlorido for (3), methanol for the second Re coordination site of (4), and the deprotonated hydroxyl of salicylidene ligand to form the dinuclear structure of (5).

The O2, C2, Re1, and Br1 atoms of complex (1) crystallized on a mirror plane resulting in half of the complex being present in the asymmetric unit where the other is generated by symmetry through the x, −y + 1, z symmetry element. Complex (3) was crystallized from a benzene-based mother liquor which resulted in one and a half benzene solvent molecules crystallizing in the asymmetric unit. The latter on a special position wherein half is in the asymmetric unit and the other generated by symmetry through the −x + 1, −y, −z symmetry element. The first benzene is disordered with a ratio of 66 : 34. The crystal structures of complexes (1) and (3) are shown in Fig. 6.

Fig. 6 Molecular diagrams of crystal structures (1) and (3) as found in the asymmetric units generated via symmetry, the atom names of (1) that are marked with asterisk indicate those generated by symmetry. The symmetry generated sections are indicated in dashed bonds and the atoms are opaquely drawn. The benzene disorder of (3) is drawn in the same manner, which has a disorder ratio of 66 : 34.

Complex (4) is disordered at the second metal site, i.e. the fac-[Re(CO)₃]⁺ core bound to the salicylidene functional group, and methanol with a disorder ratio of 51 : 49. The disorder also contains two water molecules in the asymmetric unit. Complex (5) crystallised on a special position whose full structure is generated by the -x + 1, -y + 1, -z + 1 symmetry element. It contains a water molecule and DMF solvent molecule, disordered with a 72 : 28 ratio with the major occupancy belonging to the DMF molecule. The molecular diagrams of (4) and (5) can be seen in Fig. 7.

Fig. 7 Molecular diagrams of the asymmetric unit of complexes (4) and (5). Complex (4) indicates a disorder for the rhenium-salicylidene portion of the di-metal complex with a 51 : 49 ratio. The major disorder indicated in (4) A and the minor in the (4) B with the bonds depicted as dashed lines and the atoms drawn opaquely for the minor disordered section. The full structure of complex (5) generated by symmetry is indicated. The atom names of (5) with asterisk indicate atoms generated by symmetry.

Complex (1) crystallises in C2/m and complex (5) in P2₁/c, belonging to the monoclinic Bravais lattice with four molecules in the unit cell (Z = 4). Complexes (3) and (4) has two molecules in the unit cell (Z = 2) and crystallise in the triclinic space groups, P1̄. Complex (4) proved challenging to crystallize and small needle crystals suitable for X-ray diffractions was obtained from DMSO after several weeks. The structure of (4) contains a solvent mask for 1.5 water molecules as the refinement of the water molecules proved challenging due to their proximity to the −x, −y, −x inversion centre. Hydrogen atoms bound to the water molecules (two additional molecules not contained in the solvent mask) were placed according to their idealized positions. The general X-ray crystallographic refinement parameters are listed in Table 2.

Table 2 General X-ray crystallographic and refinement parameters for (1) fac-[Re(CO)₃(Phen)(Br)]; (3) fac-[Re(CO)₃(Phen5MeSalH)(Cl)]; (4) fac-[Re(CO)₃Br(Phen5MeSal)Re(CO)₃(MeOH)]; (5) fac-[Re(CO)₃(Phen5MeSal)]₂

Compound	(1)	(3)	(4)	(5)
Quest: Bruker D8 Quest Eco Chi Photon II CPAD diffractometer; Venture: Bruker D8 Venture 4K Kappa Photon III C28 diffractometer.
Crystal data
Chemical formula	C15H8BrN2O3Re	C32H24ClN3O4Re	C27H23.04BrN3O10.52Re2	C50.31H39.19N7.43O10Re2
Mr	530.34	736.19	1010.17	1280.28
Crystal system, space group	Monoclinic, C2/m	Triclinic, P1̄	Triclinic, P1̄	Monoclinic, P21/c
Temperature (K)	101	100	109	100
Wavelength (Å)	MoKα (λ = 0.71073)	MoKα (λ = 0.71073)	CuKα (λ = 1.54178)	MoKα (λ = 0.71073)
a, b, c (Å)	15.625(1),	10.912(2),	7.817(1),	10.0739(4),
	11.8601(9),	10.982(1),	10.926(2),	21.6296(8),
	8.0033(6)	14.369(2)	18.312(3)	11.0985(4)
α, β, γ (°)	90,	110.547(5),	76.475(6),	90,
	106.752(3),	91.907(5),	84.564(8),	102.883(1),
	90	116.622(4)	81.830(7)	90
V (Å^3)	1420.2(2)	1403.6(3)	1502.1(4)	2357.4(2)
Z	4	2	2	4
µ (mm^−1)	11.383	4.467	17.623	5.198
Crystal size (mm)	0.088 × 0.065 × 0.046	0.3 × 0.038 × 0.032	0.09 × 0.023 × 0.011	0.105 × 0.058 × 0.027
Data collection
Diffractometer	Venture	Quest	Venture	Venture
Tmin, Tmax	0.550, 0.746	0.555, 0.746	0.436, 0.752	0.635, 0.746
No. of measured, independent, and observed [I > 2σ(I)] reflections	16 202, 1850, 15 117	26 522, 6553, 21 593	42 707, 4983, 35 885	58 555, 5871, 44 495
Rint	0.0482	0.0564	0.0617	0.0561
Completeness (%)	100	96.3	98.0	99.8
Refinement
R[F^2 > 2σ(F^2)], wR(F^2), S	0.0215, 0.0508, 1.119	0.0397, 0.0750, 1.183	0.0494, 0.1064, 1.082	0.0328, 0.0659, 1.080
No. of reflections	1850	6553	4983	5871
No. of parameters	100	370	429	322
Δρmax, Δρmin (e Å^−3)	1.53, −1.35	1.39, −2.23	1.53, −3.29	1.64, −2.08
CCDC No.	2172526	2424429	2424430	2424431

---

All the structures form distorted octahedral geometries with respect to the rhenium metals bound to the 1,10-phenanthroline ligand as reflected by the N1–Re–N2/1 angle which is identical within standard uncertainty with values of 76.0(2)°, 75.8(2)°, 75.7(3)°, and 76.2(1)° for (1), (3), (4), and (5). The same is true for the C1–Re1–C2 angle which is 89.5(1)°, 88.8(3)°, and 88.9(5)°. The C1–Re1–C2 angle for (5) is 86.5(2)° deviating slightly from the series. The Re1–N1 distances are all similar throughout the series being 2.170(3) Å, 2.186(4) Å, 2.16(1) Å, and 2.172(3) Å for (1), (3), (4), and (5). The Re1–O4 and Re1–O5 distances seen in compounds (4) and (5) vary in accordance to the di-metal versus dinuclear conformations. The Re1–O4 distance (to the O atom of the salicylidene hydroxyl) has values of 2.11(1) Å for (4) with extension observed for the bridging (5) of 2.138(3) Å. The Re1–O5 distance (the methanol hydroxyl atom) for (4) is 2.19(1) Å and is of similar order to Re1–O4. The steric strain of the dinuclear complex (5) relative to (4) is also observed in the O4–Re1–C2 angle being 96.9(1) versus 91.1(8) °. The remaining bonds and distances are similar to that found in literature for related crystal structures.^54–56 The bond distances and angles are listed in Table 3.

Table 3 Selected bond distance and angle parameters (Å and °) of complexes (1), (3), (4), and (5)

Complex	Re1–N1 (Å)	Re1–N2 (Å)	Re1–O4	Re1–O5	N1–Re1–N2/1 (°)	C1–Re1–C2 (°)	O4–Re1–C2 (°)	C17–C16–N3–C8/10 (°)
a Only the major occupancy is considered in these bonds and angles.
(1)	2.170(3)	N/A	—	—	76.0(2)	89.5(1)	—	-N/A
(3)	2.186(4)	2.183(5)	—	—	75.8(2)	88.8(3)	—	-175.0(5)
(4)^a	2.16(1)	2.169(7)	2.11(1)	2.19(1)	75.7(3)	88.9(5)	91.1(8)	-173.3(9)
(5)	2.172(3)	2.181(3)	2.138(3)	—	76.2(1)	86.5(2)	96.9(1)	−177.5(3)

---

The hydrogen bond interactions were analysed in detail and are listed in the SI. These interactions include both traditional hydrogen interactions (with oxygen or nitrogen acceptors) for all the structures and halogen interactions for complexes (3) and (4) and correlated well to the Hirshfeld analysis. The molecular structures were overlayed to illustrate the atomic variations induced by the formation of mononuclear, dinuclear and di-metal species in comparison to the phenanthroline bifunctional ligand (L) (Fig. 8). Ligand (L) has two crystallographic unique structures YELTIW⁴⁰ and OKOHIH⁴¹ denoted as (L) and (L′). The ligands, crystallized under various conditions incorporated different solvent molecules in the crystalline structures. The alternative space group and coordinated solvent affect the relative orientation of the molecule causing the rotational variation between the molecules (L+L′) when overlayed with a dihedral angle of 86.02°. These two relative orientations (L or L′) occur preferentially for the respective organometallic complexes as can be visualized in the overlays.

Fig. 8 Graphical overlay of combinations between the four 1,10-phenanthroline rhenium complexes and (L) as well as an overlay of (L) and (L′). The left inset is the overlay of all the ligands and complexes. Each overlay is generated in Mercury⁵⁷ by overlaying the 14-membered aromatic 1,10-phenanthroline group between the structures (N1, N2, and C4–C15). The RMS values are calculated in the same manner. RMS values are: 0.051 Å for (L+L′); 0.0569 Å for (L+1); 0.0567 Å for (L′+3); 0.0356 Å for (L+4); 0.0449 Å for (L+3); 0.0455 Å for (L+5); 0.0541 Å for (1+3); 0.0515 Å for (1+4); 0.0384 Å for (1+5); 0.0303 Å for (3+4); 0.0387 Å for (3+5) and 0.0326 Å for (4+5).

The phenanthroline functional group aligns well for (1) and (L) illustrating the steric and rotational freedom which the salicylidene functional group can adopt. The mononuclear complex (3)'s salicylidene group has crystallized in an orientation that is similar to that of (L′), whereas the dinuclear specie (5) has a similar orientation (L). The di-metal complex (4) has crystallized in an orientation markedly different from all the other structures in order to accommodate the second metal coordination to the salicylidene group. The overlay images illustrate the flexibility of the salicylidene group to rotate in crystallographic space to nearly 120° depending on immediate chemical environment.

3.3 Hirshfeld analysis

Hirshfeld surface analysis is a useful tool for determining the dominant non-covalent interactions which are observed within a small molecule crystal structure, thus indicating the prevalent interactions which may be accessible in a protein binding environment and thereby assisting the design of new pharmaceutical compounds. The Hirshfeld fingerprint plots and their respective crystal structure diagrams are shown in Fig. 9.

Fig. 9 Hirshfeld surface d-norm plots of the four crystal structures (1), (3), (4), and (5). Within each plot is indicated key interactions such as hydrogen bonding, halogen bonding and other van der Waals interactions. Below each inset is their respective crystal structures.

The fingerprint plots show H⋯H interactions for (3), (4), and (5) but not prominently observed for (1). All the structures (1, 3, 4, 5) indicate marked interactions for O⋯H which are present near to the carbonyl oxygens (M–CO) in all four structures. For complexes (3), (4), and (5) additional O⋯H interactions are observed near to the salicylidene oxygen atoms. Halogen–hydrogen interactions are observed for complexes containing the coordinated Br/Cl. For complexes (3) and (4) the halogen⋯H interactions are strong enough to be observed as formal hydrogen interactions (see SI Table S1). Complex (1) although displaying hydrogen bromide interactions, it is viewed as weak and only one formal hydrogen interaction is seen from the oxygen donor.

The Hirshfeld surface can be partitioned into different sections based on which atom pair types are interacting at the surface on a given point. Thus, it can be used to classify the relative contribution of the interactions as a function of percentage. From this classification the largest type of interaction seen on the surfaces of the metal complexes are those between oxygen and hydrogen atom pairs and H⋯H atom pairs. Complexes (1), (3), (4), and (5) has O⋯H surface percentages of 27.1%, 24.5%, 37.3%, and 28.1%. Both (3) and (5) have larger surface percentages dedicated to H⋯H type interactions (29.7% and 35.1%) than O⋯H based interactions and the contrary being true for (1) and (4) with H⋯H surface percentages of 11.5% and 18.8%. C⋯H interactions are the third largest in percentage values of 20.3%, 23.1%, 16.5%, and 18.5% for (1), (3), (4), and (5) respectively. Complexes (1), (3), and (4) has halogen⋯H interactions with percentage values of 15.7%, 8.0%, and 7.2%. The remainder of the surfaces are made up of Br/Cl⋯C, N⋯H, C⋯O, C⋯C, or O⋯O type interactions. The interactions and relative percentage contribution which constitute the Hirshfeld surface indicates a near equal preference for hydrophobic and hydrophilic regions of space (Fig. 10) but dominated by interactions directed to H atoms. A full breakdown of the Hirshfeld surfaces can be found in the SI.

Fig. 10 Hirshfeld surface percentage contribution of each atom pair type per complex.

3.4 Cytotoxicity and drug likeness testing

In silico analysis of the drug-likeness of compounds (L, 1–5) was conducted using the SwissADME web service (https://swissadme.ch/). The applicability domain of SwissADME's predictive models is restricted to druglike organic compounds, and the tool is not parameterised for metal centres or the coordination bond.⁵⁸ Consistent with metallodrug literature,^59–61 SwissADME was applied to probe the drug-likeness of the organic ligand component of each compound, and results for the metal complexes are interpreted accordingly as indicative rather than quantitatively predictive, recognising that metal coordination modifies physicochemical properties (such as log P, aqueous solubility, and membrane permeability) in ways that lie outside the platform's training data. With this caveat, the analysis provides a useful comparative baseline for assessing the relative drug-likeness across the series and identifying how structural variation in the ligand framework modulates predicted properties. The full SwissADME results are provided in the SI. For the free ligand L and complexes 1–4, high gastrointestinal absorption is predicted, with potential for blood–brain barrier penetration; compound 5 is predicted to have lower pharmacokinetic potential. CYP inhibition predictions for the ligand-based framework suggest L may inhibit CYP1A2, CYP2C19, CYP2D6, and CYP3A4; compound 1 may inhibit CYP1A2, CYP2C9, CYP2D6, and CYP3A4; compounds 2 and 3 may inhibit CYP1A2, CYP2C19, and CYP2D6; compound 4 shows predicted inhibition only for CYP2C9; and compound 5 shows no predicted inhibition. These trends suggest that structural variation across the series could enable tuning of inhibitory selectivity, though experimental confirmation will be required. Given that, in silico tools cannot substitute for direct biological evaluation of metal complexes, in vitro cytotoxicity testing against HeLa and MCF-7 cells was conducted as described below.

Further more, rhenium complexes can induce a particular cellular response depending on the atomic and structural variations incorporated by the ligands bound to the complex.⁶² Rhenium based 1,10-phenanthroline complexes have shown promise as potential anti-tumor compounds⁶³ as illustrated in the list below of respective rhenium complexes which utilized the functional group (Table 4). The specific cell lines and the proposed mechanism of action by which the compound acts against the cancer cells are listed.

Table 4 List of rhenium 1,10-phenanthroline based complexes which indicated antitumor activity, included are the cell lines used for testing as well as the mechanism of action

Compound name/acronym^a	Cell line	Mechanism of action	Ref
a Due to the complexity of some the compounds the abbreviations of the compounds are used as described in literature when possible: dmphen = 2,9-dimethyl-1,10-phenanthroline; phen = 1,10-phenanthroline; DAPTA = 1,4-diacetyl-1,3,7-triaza-5-phosphabicylco[3.3.1]nonane; phendione = 1,10-phenanthroline-5,6-dione; dip = 4,7-diphenyl-1,10-phenanthroline; ROS = reactive oxygen species.
fac-[Re(CO)3(dmphen)(OH2)]^+	A2780, A2780CP70	Growth inhibition	64
[ReO(OMe)(dip)Cl2]	A549	Necroptosis	65
[ReO(OMe)(3,4,7,8-tetramethyl-1,10-phenanthroline)Cl2]	A549	Necroptosis	65
Re-ART-1	HeLa	Apoptosis/ferroptosis	66
fac-[Re(CO)3(dmphen)(para-tolyl isonitrile)]	A2780	Apoptosis	67
fac-[Re(CO)3(TBS-pyridocarbazol)(Py)]	HeLa	Photodynamic therapy	68
fac-[Re(CO)3(phen)(DAPTA)]^+	HeLa, A2780, A2780CP70	Photodynamic therapy	69
fac-[Re(CO)3(dmphen)(pentyl carbonato)]	PC-3, MDA-MB-231, CCl-227	DNA intercalation	70
fac-[Re(CO)3(phendione)Cl]	T98G, PC3, MCF-7	DNA intercalation	71
Rhenium(I)-DCA conjugate	NCI-1229	Apoptosis	72
fac-[Re(CO)3(4-(1H-imidazo[4,5-f] [1,10]phenanthrolin-2-yl)aniline) (Imi)]^+	HepG2, HeLa, MCF-7, A549	ROS apoptosis	73
fac-[Re2(CO)6(dip)2(4,4′-azopyridine)](PF6)2	HeLa, A549, MCF-7	Apoptosis	74
fac-[Re(CO)3(dip)(Py-3-CH2Cl)]^+	A549	ROS apoptosis	75
fac-[Re(CO)3(phen)(4-(1-ferrocenyl)-butanoate)]	HeLa	ROS apoptosis	76
fac-[Re2(CO)6(dip)2(4,4′-(ethane-1,2-diyl)dipyridine)]	HeLa, A549, HepG2	Apoptosis/paraptosis	77
CA-Re	MDA-MB-231	Photodynamic and immunotherapy	78

---

Viewing each of the above 1,10-phenanthroline complexes from a structure activity relationship perspective reveals certain trends. In Fig. 11–13, the complexes have been illustrated by separating the coordinated phenanthroline (blue highlighted) and additional monodentate ligand on the metal's sixth position (green highlighted). For the purpose of this discussion any type of programmed cell death is grouped together, this includes apoptosis, ROS apoptosis, ferroptosis, and necroptosis. The bidentate ligand 4,7-diphenyl-1,10-phenanthroline (abbreviated as dip) appears to cause programmed cell death in all the complexes where it is present. These complexes are (a) [ReO(OMe)(dip)Cl₂], (b) fac-[Re₂(CO)₆(dip)₂(4,4′-azopyridine)](PF₆)₂, (c) fac-[Re(CO)₃(dip)(Py-3-CH₂Cl)]⁺, (d) fac-[Re₂(CO)₆(dip)₂(4,4′-(ethane-1,2-diyl)dipyridine)], and the (e) 2,9-dimethyl-4,7-diphenyl-1,10-phenanthroline containing rhenium(I)-DCA conjugate. These complexes have the adapted dip-ligand, with a smaller neutral ligand in the 6^th position of the octahedral sphere. In the case of (b) and (d) the neutral linking ligand is the means to form a symmetrical di-metal complex.

Fig. 11 Line drawings of the rhenium-phenanthroline based complexes which have been tested against cancer cell lines. The complexes have been grouped according to structural similarity. Blue indicates the phenanthroline-based bidentate ligand systems where green shows the ligand systems which bind as monodentate ligands to the rhenium atom.

Fig. 12 A diagram of two possible structural variations (i.e. substitution versus adaption) by which the rhenium-1,10-phenanthroline core can be manipulated for potential anti-cancer applications. The green highlighting indicates the substitution of a ligand (R³) on the sixth position on the metal centre. The blue highlighting indicates the chemical adaptation of the 1,10-phenanthroline ligand varying from least (R¹) to increasing (R²) steric interference.

Fig. 13 Line drawings of the five complexes (1–5) which have been tested against HeLa and MCF-7 cells in an SRB assay during this study. The blue section indicates the 1,10-phenantroline based ligands and the green sections highlights the sixth position of the metal complexes where monodentate variation may be found.

Other complexes which cause programmed cell death is (f) fac-[Re(CO)₃(phen)(4-(1-ferrocenyl)-butanoate)], (g) Re-ART-1, and (h) fac-[Re(CO)₃(dmphen)(para-tolyl isonitrile)]. These three complexes have the non-steric demanding phen ligand or adapted meta to the N,N′ atoms, again limiting the steric demand of the ligand. The monodentate ligands bind well to the sixth position, and known to be bio-capable (isonitriles, ferrocene, etc.) thus may be an influencing force responsible for the cytotoxic effect compared to the other complexes which has the same bidentate phenanthroline ligand system (dmphen or phen) that either only inhibits growth ((i) fac-[Re(CO)₃(dmphen)(OH₂)]⁺) or was used in photodynamic therapy ((j) fac-[Re(CO)₃(phen)(DAPTA)]⁺).

The (k) fac-[Re(CO)₃(4-(1H-imidazo[4,5-f][1,10]phenanthrolin-2-yl)aniline)(Imi)]⁺ complex also causes programmed cell death, however what is more revealing is that its adapted phen ligand, seems structurally to fit photodynamic therapy and is sterically more demanding. The other complexes with similar bidentate ligand systems (containing a five-membered heterocycle bound to the phenanthroline) are (l) fac-[Re(CO)₃(TBS-pyridocarbazol)(Py)] and (m) CA-Re, both of which are tested successfully for photodynamic therapy. Complexes (n) fac-[Re(CO)₃(dmphen)(pentyl carbonato)] and (o) fac-[Re(CO)₃(phendione)Cl] both induces cellular death by DNA intercalation despite having only slight modifications to the phenanthroline bidentate ligand when compared to (h) fac-[Re(CO)₃(dmphen)(para-tolylisonitrile)] and (p) [ReO(OMe)(3,4,7,8-tetramethyl-1,10-phenanthroline)Cl₂] (containing smaller ligands in the 6^th position) where programmed cell death was the result. Line drawings of the complexes, grouped according to structural similarity and respective functional groups highlighted, are illustrated in Fig. 11.

By studying the structural differences between these complexes two distinct variations can be observed. One is the substituent on the sixth position on the metal complex whether it is tightly bound, sterically undemanding or bio-capable and large in its own right. Skiba et al.⁷⁶ has shown that the complex fac-[Re(CO)₃(1,10-phenanthroline)(4-(1-ferrocenyl)-butanoate)] can induce upregulation of ROS adducts forcing apoptosis. Marker et al.⁶⁹ used the DAPTA monodentate ligand to facilitate cell death in HeLa, A2780, and A2780CP70 cell lines with the fac-[Re(CO)₃(phen)(DAPTA)]⁺ complex in photodynamic therapy. The second structural variation is the adaptation of the 1,10-phenanthroline ligand itself by replacing hydrogen atoms with functional groups, meta to the N,N′ atoms (least steric demand), ortho or para thus increasing spatial requirements. Growth inhibition of A2780 and A2780CP70 cells could be demonstrated by Konkankit et al.⁶⁴ when using dmphen (2,9-dimethyl-1,10-phenanthroline), a methylated version of 1,10-phenanthroline in the fac-[Re(CO)₃(dmphen)(OH₂)]⁺ complex. Alternatively, a combination of these variations can be used. A dimeric complex [Re₂(CO)₆(dip)₂(4,4′-(ethane-1,2-diyl)dipyridine)] (where dip = 4,7-diphenyl-1,10-phenanthroline) has been shown to cause apoptosis and paraptosis in HeLa, A549, and HepG2 cells by Ye et al.⁷⁷ A schematic representation of the two adaptation ideologies is shown in Fig. 12.

In our study the rhenium-1,10-phenanthroline based complexes (1–5) were systematically adapted both by functionalization of the 1,10-phenanthroline ligand and while ensuring that the ligand on the sixth position is occupied by a monodentate halide or solvento specie and therefore substitution to form protein-metal coordination can take place next to the metal center. Preliminary biological screening to test their cytotoxicity against HeLa and MCF-7 cells were conducted. Line drawings of each of these complexes with the monodentate ligands (highlighted in green) and bidentate ligands of the 1,10-phenanthroline-based ligand (indicated in blue) are shown in Fig. 13.

The complexes were tested against HeLa and MCF-7 cells by means of an SRB assay to obtain their IC₅₀ values to measure cell death as a function of compound concentration, i.e. the concentration (in μM) needed to inhibit cell growth by 50% in vitro, thus the lower the value, the more cytotoxic the compound is. The complexes were tested against cisplatin as the internal standard which gave an IC₅₀ value of 3 ± 3 μM and 23 ± 2 for HeLa and MCF-7 respectively. A graph of the IC₅₀ values are shown in Fig. 14. The IC₅₀ values for HeLa of all compounds are higher than that of cisplatin but still shows notable cytotoxic activity. Cytotoxic testing for HeLa of the aqua complex of (1) (i.e. fac-[Re(Phen)(CO)₃(H₂O)]) has been conducted by Knopf et al.⁷⁹ reporting an IC₅₀ value 9.7 ± 5.8 μM, lower activity than the 26 ± 3 μM observed for (1). The cytotoxic work done by Knopf et al. was according to a MTT assay and is therefore not directly comparable. However, it should be noted that the coordinated bromide on the sixth position of the metal complex would affect activity, as it is likely that the aqua complex is more soluble and would substitute the bound water more readily than the electronically attractive Br ion.

Fig. 14 Bar graph depicting the IC₅₀ values of HeLa and MCF-7 cells with respect to all compounds relative to the cisplatin standard. Error bars are indicated above each entry. The compounds were incubated with the cells for 3 days at 37 °C.

The functionalized ligand (L) displays mildly more cytotoxic activity than pure phenanthroline (IC₅₀ values 9 ± 3 μM versus 11 ± 2 μM for HeLa) whereas its mildly less cytotoxic for MCF-7 (6 ± 3 μM versus 8 ± 2 μM). Upon coordination to rhenium, there is a general reduction of toxicity between the activity of the free ligands (phen or L) versus the metal coordinated species (1–5) for both cell lines.

Coordination of the phen to the rhenium bromide specie (1) decreases the HeLa cytotoxicity activity with IC₅₀ of 11 ± 2 μM versus 26 ± 3 μM. The reduction is 18% less pronounced for the functionalized ligand (L) (9 ± 3 μM) relative to complex (2) and (3), which display similar and moderate cytotoxicity with IC₅₀ values of 15 ± 3 μM, 16 ± 3 μM, containing bromide versus chloride in the 6^th position. The asymmetrical di-metal complex (4) containing two metal centers, (bound to phenanthroline and salicylidene) has poorer cytotoxic (IC₅₀ value of 22 ± 2 μM) relative to (2)/(3) and is similar to (1). The symmetrical dinuclear complex (5) is the least cytotoxic of the series with an IC₅₀ value of 35 ± 3 μM. A number of factors may affect the decreased activity of (5), such as it having the lowest solubility of the series; it is sterically rigid with all coordination sites fully occupied by bound atoms thus preventing any metal-protein covalent binding. It also shows the highest percentage contribution of H⋯H interactions (35.1%) i.e. more lipophilic character, with low C⋯C interactions associated with π–π interactions. Complex (2)/(3) indicating the best cytotoxicity of the Re coordinated complexes have the coordinated halido atom in the sixth position, which can be readily substituted to promote Re-protein covalent binding. Being mononuclear complexes, their spatial occupation would be the least, while (3) appears to have dominant C⋯H interactions (23.2%) moderate O⋯H, H⋯H, C⋯C and low C⋯O (2.4%) interactions. It is notable too that (3) has distinct face-to-face π–π interactions with the coordinated benzene molecule (Fig. SI 22). The findings suggests that optimal anticancer activity requires balanced amphiphilic character approaching a Goldilocks “just right” principal: sufficient C⋯H interactions for protein recognition, moderate O⋯H bonding for aqueous solubility, and balanced H⋯H contacts for membrane permeability. The activity of (3) indicates that C–H⋯π interactions—which enable favourable aromatic stacking with protein residues may be more beneficial for cytotoxicity than purely hydrophobic (H⋯H) or purely polar (O⋯H) interactions.

In comparison to the HeLa assay, all metal complexes show reduced cytotoxicity for MCF-7. Complex (4) with its intrinsic flexibility and functionalisation possible on the metal solvent/halido coordination sites, has scope for optimisation in order to increase the toxicity. Complex (1), (2) and (5) show negligible toxicity for MCF-7, while (L) and phen are more cytotoxic than the cisplatin standard.

Organic Schiff-bases has shown moderate selectivity for cancer cells (MCF-7, HeLa, and A549 vs. Vero cells).⁸⁰ The ligand (L) had demonstrated good selectivity against cancer cells (A549, HepG2, 4T1, DU145, and Caco-2) when measured against Vero cells.⁸¹ Enslin et al. synthesised a series of functionalised rhenium(I)tricarbonyl-1,10-phenantroline complexes and studied the selectivity of the complexes against PC-3 cancer cells vs. normal RPE-1 cells and found that the compounds were more selective for the cancer cells.⁸² Parson et al. studied a series of pentylcarbonato bound rhenium(I)tricarbonyl complexes (with modifications to the 1,10-phenantroline bidentate ligand) against lymphosarcoma, PC-3, and myeloid leukemia cancers cells. The findings showed great cytotoxicity and selectivity for the cancer cells when compared to normal mesangial cells.⁸³ The literature suggests that functionalized rhenium(I)tricarbonyl-1,10-phenantroline compounds may be more selective for cancer cells than normal cells.

Although the complexes indicate lower cytotoxicity to (L) and cisplatin standard particularly for HeLa cells, they are envisioned for future use in radioactive therapy which can have lower chemical cytotoxicity as it will be accompanied by the metal center's radionuclide properties to induce cell death. Therefore, complexes of lower chemical cytotoxicity can be beneficial⁸⁴ particularly if a radio-imaging agent is envisioned to avoid unnecessary radio-therapeutic dosages. The advantage of complexes (4) and (5) is the presence of two metal centers per molecule, implying their potential use as theranostic compounds, with one metal that can be used as an imaging agent (e.g. ^99mTc) and the other a therapeutic isotope such as ^188/186Re.

4 Conclusion

The development of engineered chemical compounds is needed to combat the ever-rising number of global cancer occurrences and drug resistance. New chemical scaffolds may prove useful, with metallodrug complexes indicated as promising candidates due of their variable chemical, redox, and physical properties not possible for purely organic molecules containing on s and p electronic orbitals. To this end, a series of modified 1,10-phenantroline based rhenium tricarbonyl complexes were developed, synthesized and tested against HeLa and MCF-7 cells. Variation in the synthetic procedure of the metal complexes was explored using various metal precursors. The use of [Re(CO)₅X] (where X = Cl, Br) instead of fac-2Et₄N·[Re(CO)₃Br₃] (ReAA) resulted in the successful crystallization of the complexes which were not achievable when using ReAA, as well as higher yields were obtained without the need for additional purification steps to ensure the removal of [Et₄N]Br salt. The controlled formation of mononuclear, dinuclear and mixed di-metal species were possible by engineering the synthetic procedure.

Structural elucidation of the metal complexes was achieved using SC-XRD. In-depth solid-state structural investigations through the use of Hirsfeld surface analysis revealed that halogen⋯H, H⋯O, and H⋯H type interactions are the most prominent weak interactions observed. Consideration was noted for possible covalent and non-covalent interactions which the metal complex may be prone to with respect to future biological environments. The in silico drug likeness of the compounds were analysed and had shown that all but compound (5) is predicted to pass the blood brain barrier and that the structural changes between the compounds in the series had changed the predicted drug target inhibition for the compounds (L, 1–5). The structural activity relationships were investigated for HeLa and MCF-7 cell toxicity results. The ligand and complexes have low toxicity for MCF-7, but show preferential toxicity for HeLa cell with IC₅₀ values of 9 ± 3 μM, 26 ± 3 μM, 15 ± 3 μM, 16 ± 3 μM, 22 ± 2 μM, and 35 ± 3 μM for (L), (1)–(5) respectively with best metallo-induced cytotoxicity being indicated for (2) and (3) which show a balanced distribution of weak interactions accompanied by π–π interactions. The metal complexes have potential scope for improved therapy with the application of an appropriate radionuclide (i.e. ¹⁸⁶Re and/or ¹⁸⁸Re) which would support the cellular death of a diseased cell and dependence would not rely solely on chemo toxic properties of the compound in question. The inclusion of imaging radionuclide (such as ^99mTc) in the dinuclear or di-metal specie would allow for the development of theranostic compounds.

Author contributions

Frederick J. F. Jacobs: writing – original draft, visualization, validation, data curation, investigation, methodology; Eleanor Fourie: resources, formal analysis; Alice Brink: conceptualization, data curation, writing – review & editing, resources, funding acquisition, project administration, supervision. The manuscript was written through contributions of all authors. All authors have given approval to the final version of the manuscript.

Conflicts of interest

There is no conflicts of interest to report.

Data availability

The data supporting this article have been included as part of the supplementary information (SI). Supplementary information is available and includes spectroscopic, crystallographic, cytotoxic, and in-silico results pertaining to the manuscript. See DOI: https://doi.org/10.1039/d6dt00636a.

CCDC 2172526 (1), 2424429 (3), 2424430 (4) and 2424431 (5) contain the supplementary crystallographic data for this paper.^85a–d

Acknowledgements

This research is supported by the South African National Research Foundation (Grant No.: 137759; NEP grant: 116180). AB & FJFJ thank J.R. Helliwell (University of Manchester, UK) for his ongoing support in macromolecular crystallography and synchrotron development. Drs M. R. Swart and C. Marais (University of the Free State, SA) are thanked for NMR guidance.

References

1. G. Ebers, Papyrus Ebers The Hermetic book of The Medicines of Old Egypt in Hieratic Font, Wilhelm Engelmann, Liepzig, 1875.
2. M. Arruebo, N. Vilaboa, B. Sáez-Gutierrez, J. Lambea, A. Tres, M. Valladares and Á. González-Fernández, Cancers, 2011, 3, 3279–3330.
3. U. Anand, A. Dey, A. K. S. Chandel, R. Sanyal, A. Mishra, D. K. Pandey, V. De Falco, A. Upadhyay, R. Kandimalla, A. Chaudhary, J. K. Dhanjal, S. Dewanjee, J. Vallamkondu and J. M. Pérez de la Lastra, Genes Dis., 2023, 10, 1367–1401.
4. M. Kciuk, A. Gielecińska, S. Mujwar, D. Kołat, Ż. Kałuzińska-Kołat, I. Celik and R. Kontek, Cells, 2023, 12, 659–689.
5. S. Ghosh, Bioorg. Chem., 2019, 88, 102925–102945.
6. A. Brink, F. J. F. Jacobs and J. R. Helliwell, IUCrJ, 2022, 9, 180–193.
7. B. R. Crane, A. J. Di Bilio, J. R. Winkler and H. B. Gray, J. Am. Chem. Soc., 2001, 123, 11623–11631.
8. C. Shih, K. A. Museth, M. Abrahamsson, A. M. Blanco-Rodriguez, A. J. Di Bilio, J. Sudhamsu, B. R. Crane, K. L. Ronayne, M. Towrie, A. J. Vlček, J. H. Richards, J. R. Winkler and H. B. Gray, Science, 2008, 320, 1760–1762.
9. H. Duan, A. Iagaru and C. M. Aparici, Nanotheranostics, 2022, 6, 103–117.
10. J. Koziorowski and J. Ballinger, Q. J. Nucl. Med. Mol. Imaging, 2022, 65, 306–314.
11. H. Y. Tan, C. H. Yeong, Y. H. Wong, M. McKenzie, A. Kasbollah, M. N. Md. Shah and A. C. Perkins, Nucl. Med. Biol., 2020, 90–91, 55–68.
12. C.-H. Yeong, Med. Phys. Int., 2022, 10, 363–370.
13. P. V. Harper, K. A. Lathrop, F. Jiminez, R. Fink and A. Gottschalk, Radiology, 1965, 85, 101–109.
14. D. Papagiannopoulou and J. Label, Compd. Radiopharm., 2017, 60, 502–520.
15. M. Riondato, D. Rigamonti, P. Martini, C. Cittanti, A. Boschi, L. Urso and L. Uccelli, J. Med. Chem., 2023, 66, 4532–4547.
16. L. Uccelli, P. Martini, L. Urso, T. Ghirardi, L. Marvelli, C. Cittanti, A. Carnevale, M. Giganti, M. Bartolomei and A. Boschi, Molecules, 2022, 27, 5283.
17. S. M. Qaim, Nucl. Med. Biol., 2017, 44, 31–49.
18. M. Argyrou, A. Valassi, M. Andreou and M. Lyra, Int. J. Mol. Imaging, 2013, 2013, 1–7.
19. G. Koca, H. Ozsoy, H. I. Atilgan, K. Demirel, V. E. Dincel and M. Korkmaz, Nucl. Med. Mol. Imaging, 2012, 46, 215–217.
20. N. Ö. Küçük, E. İbiş, G. Aras, S. Baltaci, G. Özalp, Y. Bedük, N. Çanakci and A. Soylu, Ann. Nucl. Med., 2000, 14, 239.
21. H.-C. Ni, C.-Y. Yu, S.-J. Chen, L.-C. Chen, C.-H. Lin, W.-C. Lee, C.-H. Chuang, C.-L. Ho, C.-H. Chang and T.-W. Lee, Appl. Radiat. Isot., 2015, 99, 117–121.
22. B. Lambert, K. Bacher, L. Defreyne, F. Gemmel, H. Van Vlierberghe, J. M. Jeong, R. A. Dierckx, C. Van de Wiele, H. Thierens and F. De Vos, J. Nucl. Med., 2005, 46, 60–66.
23. A. Kumar, D. N. Srivastava, T. T. M. Chau, H. D. Long, C. Bal, P. Chandra, L. T. Chien, N. Van Hoa, S. Thulkar, S. Sharma, L. H. Tam, T. Q. Xuan, N. X. Canh, G. S. Pant and G. P. Bandopadhyaya, Radiology, 2007, 243, 509–519.
24. S. Baxi, S. Vohra, A. Hong, N. Mulholland, M. Heuschkel, G. Dahlhoff, G. Cardaci, S. Mirzaei and M. Sathekge, J. Nucl. Med., 2024, 65, 1450–1455.
25. J. D. McKinney, Toxicol. Sci., 2000, 56, 8–17.
26. B. L. Murphy, S. C. Marker, V. J. Lambert, J. J. Woods, S. N. MacMillan and J. J. Wilson, J. Organomet. Chem., 2020, 907, 121064–121072.
27. H. C. Bertrand, S. Clède, R. Guillot, F. Lambert and C. Policar, Inorg. Chem., 2014, 53, 6204–6223.
28. D. Braga and F. Grepioni, Acc. Chem. Res., 1997, 30, 81–87.
29. K. Takematsu, H. R. Williamson, P. Nikolovski, J. T. Kaiser, Y. Sheng, P. Pospíšil, M. Towrie, J. Heyda, D. Hollas, S. Záliš, H. B. Gray, A. Vlček and J. R. Winkler, ACS Cent. Sci., 2019, 5, 192–200.
30. H.-P.-T. Ngo, D. Q. Nguyen, H. Park, Y. S. Park, K. Kwak, T. Kim, J. H. Lee, K. S. Cho and L.-W. Kang, BMB Rep., 2022, 55, 439–446.
31. E. Ritter, P. Przybylski, B. Brzezinski and F. Bartl, Curr. Org. Chem., 2009, 13, 241–249.
32. A. Kajal, S. Bala, S. Kamboj, N. Sharma and V. Saini, J. Catal., 2013, 2013, 1–14.
33. T. Manna, S. Maji, M. Maity, B. Debnath, S. Panda, S. A. Khan, R. Nath and M. J. Akhtar, Mol. Diversity, 2025, 29, 817–848.
34. L. Alfonso, G. Ai, R. C. Spitale and G. J. Bhat, Br. J. Cancer, 2014, 111, 61–67.
35. P. G. Sammes and G. Yahioglu, Chem. Soc. Rev., 1994, 23, 327–334.
36. L. Riccardi, V. Genna and M. De Vivo, Nat. Rev. Chem., 2018, 2, 100–112.
37. S. R. Thomas and A. Casini, Curr. Opin. Chem. Biol., 2020, 55, 103–110.
38. J. Reedijk and P. H. M. Lohman, Pharm. Weekbl., 1985, 7, 173–180.
39. A. Frei, P. P. Mokolokolo, R. Bolliger, H. Braband, M. S. Tsosane, A. Brink, A. Roodt and R. Alberto, Chem. – Eur. J., 2018, 24, 10397–10402.
40. F. J. F. Jacobs, E. Fourie, E. M. Njoya, T. J. Makhafola and A. Brink, J. Mol. Struct., 2026, 1355, 144973–144989.
41. M. S. Tsosane, H. G. Visser and A. Brink, Z. Kristallogr. - New Cryst. Struct., 2016, 231, 237–239.
42. R. Alberto, R. Schibli and P. A. Schubiger, Polyhedron, 1996, 15, 1079–1089.
43. Bruker, COSMO, Version 1.48, Bruker AXS Inc., Madison, Wisconsin U.S.A., 2003.
44. Bruker, SAINT-PLUS, Version 7.12 (including XPREP), Bruker AXS Inc., Madison, Wisconsin, USA, 2004.
45. Bruker, SADABS, Version 2016/2, Bruker AXS Inc., Madison, Wisconsin U.S.A., 2016.
46. G. M. Sheldrick, Acta Crystallogr., Sect. A, 2015, 71, 3–8.
47. L. J. Farrugia, J. Appl. Crystallogr., 2012, 45, 849–854.
48. O. V. Dolomanov, L. J. Bourhis, R. J. Gildea, J. A. K. Howard and H. Puschmann, J. Appl. Crystallogr., 2009, 42, 339–341.
49. K. Brandenburg and H. Putz, Crystal Impact GbR, Postfach, 2005, 1251, 53002.
50. V. Vichai and K. Kirtikara, Nat. Protoc., 2006, 1, 1112–1116.
51. M. A. Spackman and D. Jayatilaka, CrystEngComm, 2009, 11, 19–32.
52. M. A. Spackman and P. G. Byrom, Chem. Phys. Lett., 1997, 267, 215–220.
53. P. R. Spackman, M. J. Turner, J. J. McKinnon, S. K. Wolff, D. J. Grimwood, D. Jayatilaka and M. A. Spackman, J. Appl. Crystallogr., 2021, 54, 1006–1011.
54. H. D. Batey, A. C. Whitwood and A.-K. Duhme-Klair, Inorg. Chem., 2007, 46, 6516–6528.
55. A. Świtlicka-Olszewska, T. Klemens, I. Nawrot, B. Machura and R. Kruszynski, J. Lumin., 2016, 171, 166–175.
56. T. Cardinaels, J. Ramaekers, P. Nockemann, K. Driesen, K. Van Hecke, L. Van Meervelt, S. Lei, S. De Feyter, D. Guillon, B. Donnio and K. Binnemans, Chem. Mater., 2008, 20, 1278–1291.
57. C. F. MacRae, I. Sovago, S. J. Cottrell, P. T. A. Galek, P. McCabe, E. Pidcock, M. Platings, G. P. Shields, J. S. Stevens, M. Towler and P. A. Wood, J. Appl. Crystallogr., 2020, 53, 226–235.
58. A. Daina, O. Michielin and V. Zoete, Sci. Rep., 2017, 7, 42717.
59. M. Behzad, L. Ghasemi, M. Dusek and M. Kučeráková, Appl. Organomet. Chem., 2025, 39, e7918.
60. A. M. Abbas, H. H. Nasrallah, A. Aboelmagd, W. C. Boyd, H. Kalil and A. S. Orabi, Inorganics, 2025, 13, 144.
61. J. Sączewski, Ł. Popenda and J. Fedorowicz, Appl. Sci., 2024, 14, 8865.
62. P. Collery, D. Desmaele and V. Vijaykumar, Curr. Pharm. Des., 2019, 25, 3306–3322.
63. Q. Qi, Q. Wang, Y. Li, D. Z. Silva, M. E. L. Ruiz, R. Ouyang, B. Liu and Y. Miao, Molecules, 2023, 28, 2733.
64. C. C. Konkankit, A. P. King, K. M. Knopf, T. L. Southard and J. J. Wilson, ACS Med. Chem. Lett., 2019, 10, 822–827.
65. K. Suntharalingam, S. G. Awuah, P. M. Bruno, T. C. Johnstone, F. Wang, W. Lin, Y.-R. Zheng, J. E. Page, M. T. Hemann and S. J. Lippard, J. Am. Chem. Soc., 2015, 137, 2967–2974.
66. R.-R. Ye, B.-C. Chen, J.-J. Lu, X.-R. Ma and R.-T. Li, J. Inorg. Biochem., 2021, 223, 111537.
67. A. P. King, S. C. Marker, R. V. Swanda, J. J. Woods, S. Qian and J. J. Wilson, Chem. – Eur. J., 2019, 25, 9206–9210.
68. A. Kastl, S. Dieckmann, K. Wähler, T. Völker, L. Kastl, A. L. Merkel, A. Vultur, B. Shannan, K. Harms, M. Ocker, W. J. Parak, M. Herlyn and E. Meggers, ChemMedChem, 2013, 8, 924–927.
69. S. C. Marker, S. N. MacMillan, W. R. Zipfel, Z. Li, P. C. Ford and J. J. Wilson, Inorg. Chem., 2018, 57, 1311–1331.
70. J. Medley, G. Payne, H. N. Banerjee, D. Giri, A. Winstead, J. M. Wachira, J. A. Krause, R. Shaw, S. K. Pramanik and S. K. Mandal, Mol. Cell. Biochem., 2015, 398, 21–30.
71. M. Kaplanis, G. Stamatakis, V. D. Papakonstantinou, M. Paravatou-Petsotas, C. A. Demopoulos and C. A. Mitsopoulou, J. Inorg. Biochem., 2014, 135, 1–9.
72. J. Yang, Q. Cao, H. Zhang, L. Hao, D. Zhou, Z. Gan, Z. Li, Y.-X. Tong, L.-N. Ji and Z.-W. Mao, Biomaterials, 2018, 176, 94–105.
73. S.-F. He, N.-L. Pan, B.-B. Chen, J.-X. Liao, M. Huang, H.-J. Qiu, D.-C. Jiang, J.-J. Wang, J.-X. Chen and J. Sun, J. Biol. Inorg. Chem., 2020, 25, 1107–1116.
74. F.-X. Wang, J.-H. Liang, H. Zhang, Z.-H. Wang, Q. Wan, C.-P. Tan, L.-N. Ji and Z.-W. Mao, ACS Appl. Mater. Interfaces, 2019, 11, 13123–13133.
75. J. Yang, J.-X. Zhao, Q. Cao, L. Hao, D. Zhou, Z. Gan, L.-N. Ji and Z.-W. Mao, ACS Appl. Mater. Interfaces, 2017, 9, 13900–13912.
76. J. Skiba, T. Bernaś, D. Trzybiński, K. Woźniak, G. Ferraro, D. Marasco, A. Merlino, M. Shafikov, R. Czerwieniec and K. Kowalski, Molecules, 2017, 22, 809.
77. R. Ye, C. Tan, M. Chen, L. Hao, L. Ji and Z. Mao, Chem. – Eur. J., 2016, 22, 7800–7809.
78. X. Su, W. Wang, Q. Cao, H. Zhang, B. Liu, Y. Ling, X. Zhou and Z. Mao, Angew. Chem., 2022, 61, e202115800–e202115808.
79. K. M. Knopf, B. L. Murphy, S. N. MacMillan, J. M. Baskin, M. P. Barr, E. Boros and J. J. Wilson, J. Am. Chem. Soc., 2017, 139, 14302–14314.
80. H. van Dyk, F. J. F. Jacobs, R. E. Kroon, T. J. Makhafola and A. Brink, J. Mol. Struct., 2023, 1276, 134737.
81. F. J. F. Jacobs, E. Fourie, E. M. Njoya, T. J. Makhafola and A. Brink, J. Mol. Struct., 2026, 1355, 144973.
82. L. E. Enslin, K. Purkait, M. D. Pozza, B. Saubamea, P. Mesdom, H. G. Visser, G. Gasser and M. Schutte-Smith, Inorg. Chem., 2023, 62, 12237–12251.
83. C. Parson, V. Smith, C. Krauss, H. Banerjee, C. Reilly, J. Krause, J. Wachira, D. Giri, A. Winstead and S. Mandal, J. Bioproces. Biotech., 2013, 4, 1–5.
84. F. J. F. Jacobs, G. J. S. Venter, E. Fourie, R. E. Kroon and A. Brink, RSC Adv., 2021, 11, 24443–24455.
85. (a) CCDC 2172526: Experimental Crystal Structure Determination, 2026,  DOI:10.5517/ccdc.csd.cc2bxphg; (b) CCDC 2424429: Experimental Crystal Structure Determination, 2026,  DOI:10.5517/ccdc.csd.cc2mctd7; (c) CCDC 2424430: Experimental Crystal Structure Determination, 2026,  DOI:10.5517/ccdc.csd.cc2mctf8; (d) CCDC 2424431: Experimental Crystal Structure Determination, 2026,  DOI:10.5517/ccdc.csd.cc2mctg9.

---