Photopharmacology of Quinoline and Benzimidazole Azobenzene-Based Photoswitchable β‑Hematin Inhibitors

Abstract

Recent advances in antimicrobial drug design have incorporated photopharmacology to control drug behavior through light activation, alongside traditional phenotypical strategies including molecular hybridization and metal incorporation. The quinoline and benzimidazole scaffolds are important chemical frameworks identified by the Medicines for Malaria Venture and are known to exert their antiplasmodial activity by inhibiting hemozoin formation, an important target in antimalarial drug design. These scaffolds form part of growing libraries in drug discovery, aimed at developing novel and potent antimalarial agents through structural modifications that improve pharmacological profiles and overcome drug resistance. Consequently, in this study, azobenzene photoswitchable chemical entities have been integrated with quinoline and benzimidazole scaffolds through rational design, aiming to investigate their interactions with hemozoin. The photoswitchable inhibitors showed excellent photo fatigue resistance and thermal relaxation rates exceeding 13 h at 37 °C. In vitro analysis against chloroquine-sensitive NF54 and multidrug-resistant K1 Plasmodium falciparum strains revealed low micromolar activity against the K1 strain. Most notably, this study culminated in two benzimidazole-containing compounds that exhibited differing levels of β-hematin inhibition upon light exposure. Density functional theory calculations and molecular docking studies provided mechanistic insights into how photoisomerization modulates molecular reactivity and hemozoin binding affinity, highlighting the potential of these compounds to be used as light-activated antiplasmodials.

Introduction

Malaria, caused by an intraerythrocytic Plasmodium parasite, was responsible for an estimated 597 000 deaths globally in 2023 alone, with approximately 263 million cases reported that same year. Despite the rollout of the RTS,S/AS01 malaria vaccine and artemisinin-based combination therapy (ACT), the prevention and treatment of this disease remains hindered by the emergence of P. falciparum strains resistant to clinically used antimalarial drugs, threatening drugs currently within the developmental pipeline. , The recommended chemotherapeutic treatment for malaria today involves ACT, and frequently, the partner drugs for these treatment regimens are quinoline-based. The quinoline scaffold is a crucial structural component in pharmacologically active compounds, having proven efficacious in antimalarial chemotherapy, with chloroquine (CQ) being the most widely used and effective antimalarial agent. Another scaffold, which has gained increasing attention, is the benzimidazole entity, which has not only been shown to display antimalarial activity, − but also efficacy against a range of diseases. − More notably, these pharmacologically superior scaffolds are considered essential components of the Medicines for Malaria Venture (MMV) toolbox of scaffolds known to target hemozoin (Hz) formation, a critical target in antimalarial drug development. −

During the intraerythrocytic stage of the parasite’s life cycle, the parasite degrades the host’s hemoglobin. − A byproduct of this degradation process, which occurs in the acidic digestive vacuole (DV) of the parasite, is heme, which is toxic to the parasite. , To circumvent the adverse effects associated with a buildup of heme, the free heme is converted into hemozoin, an inert biocrystalline material that is nontoxic to the parasite, although toxic to humans when released from ruptured erythrocytes. , Many researchers have exploited this by developing antimalarials that interfere with this unique heme detoxification mechanism, , the most promising of which has been CQ. , It is also worth noting that parasite resistance to scaffolds that are known to target this pathway arises from mutations in the DV membrane transporter protein, P. falciparum chloroquine resistance transporter (PfCRT), not from changes in the heme detoxification pathway. Consequently, Hz formation inhibition is still considered a primary target within the drug development process of antiplasmodial agents, as it is unique to the parasite.

Due to the concerning rise in drug-resistant parasite and bacterial strains, scientists have directed their efforts toward approaches that can alleviate the selective pressures on bacteria and parasites, potentially leading to long-term or permanent eradication of resistance development. − Photopharmacology involves the incorporation of a photoswitch into the framework of a bioactive compound, in which the pharmacological activity can be controlled by light. In principle, light can activate the photoswitchable drug to do its desired job, following which the drug will “auto-deactivate” or thermally relax to the inactive form. This will reduce off-target effects and the accumulation of the active drug in the environment. A plethora of studies have focused on the synthesis of photoswitchable antibiotics, where known antibiotics are covalently linked to photoswitchable scaffolds, such as azobenzenes. − Much of the reported work has either shown minimal change or a slight loss in activity following the light-induced conformational change, ,, with few showing promise upon light irradiation. , Nonetheless, due to the light-controllable activity and spatial regulation, these drugs offer an opportunity to reduce the selective pressure on bacteria and parasites by limiting unnecessary/prolonged exposure to the active drug and allowing for a more targeted approach.

Despite extensive studies on photoswitchable antibiotics, to the best of our knowledge, no research has focused on the photoswitchable activity of (metal-based) small molecules and their pharmacological effects in the context of malaria. More specifically, this study examines the photoswitchable properties and impact of the resulting configurational change on the ability of the compounds to inhibit β-hematin (synthetic hemozoin) formation. Herein, we describe a series of photoswitchable quinoline- and benzimidazole-based compounds, designed through docking and computational studies, that offer insight into the effects of the (light-induced) configurational change on the inhibition of β-hematin formation. Extensive investigations into their photophysical properties and in vitro antiplasmodial activity against a drug-sensitive (NF54) and multidrug-resistant (K1) strain of P. falciparum were conducted. Finally, the inhibitory activity of the compounds on β-hematin formation, both in the absence and presence of light irradiation, was examined to establish any correlation with the in silico studies.

Results and Discussion

Density Functional Theory Calculations and Molecular Docking Simulations

Benzimidazole and quinoline pharmacophores were selected as the core structures for a series of azobenzene derivatives (Figure ). The geometries of these compounds, in both cis and trans conformations, were fully optimized in the gas phase using Density Functional Theory (DFT) at the B3LYP/def2-SVP level of theory. Single-point energies were subsequently calculated using the larger def2-TZVPP basis set for all atoms to improve energetic accuracy. Solvent effects were incorporated only for the frontier molecular orbital (HOMO–LUMO) analysis, which was carried out using the SMD implicit solvation model with DMSO as the solvent. The relative energies of the cis and trans isomers are summarized in Table S1 (Supporting Information), with the trans isomers being thermodynamically favored by more than 13.2 kcal/mol, consistent with previous reports. The increased thermodynamic stability (lower energy) of the trans forms is attributed to their extended π-conjugation and planar geometry.

1.

Chemical structure of the benzimidazole and quinoline pharmacophores (left), compounds containing these scaffolds known to inhibit hemozoin formation (middle), , and the general structure of the pharmacophoric-azobenzene compounds designed in this study (right).

Frontier Molecular Orbital (FMO) analysis was conducted to evaluate the electronic properties of the designed compounds. The HOMO–LUMO energy gaps (ΔE _HL), which correlate with molecular stability and chemical reactivity, were computed and are presented in Table . These values range from 2.84 to 3.67 eV, consistent with typical n → π* and π → π* transitions of azobenzene derivatives. These gaps are also similar to those associated with the red-light photoisomerization of related azobenzenes studied for potential therapeutic applications. The trans isomers display lower HOMO and LUMO energies relative to their cis counterparts (Table and Figure ), which may be attributed to enhanced π-electron delocalization resulting from increased planarity, as illustrated by the density isosurface plots. Notably, benzimidazole-based compounds 1 and 3 exhibit the largest ΔE _HL values, suggesting greater thermal stability and reduced chemical reactivity toward biomolecular targets.

1. Calculated HOMO-LUMO Energy Gaps of the cis and trans Isomers.

	calculated HOMO–LUMO values
	trans		cis
cmpd	HOMO, LUMO (eV)	HOMO–LUMO (ΔeV)	HOMO, LUMO (eV)	HOMO–LUMO (ΔeV)
1	–6.09, −2.56	3.53	–5.93, −2.27	3.66
2	–5.62, −2.58	3.03	–5.59, −2.38	3.21
3	–6.09, −2.54	3.55	–5.92, −2.25	3.67
Rh-3	–5.43, −2.59	2.84	–5.40, −2.36	3.03
5	–5.74, −2.60	3.15	–5.68, −2.35	3.32

2.

HOMO and LUMO orbitals and energy differences for compounds 1 and 5 modeled at the B3LYP/def2-TZVPP level of theory.

To further quantify molecular reactivity and potential biological activity, global chemical reactivity parameters including Ionization Potential (IP), Electron Affinity (EA), Chemical Hardness (η), Chemical Potential (μ), Electronegativity (χ), and the Electrophilicity Index (ω) were computed. These descriptors, calculated from frontier orbital energies (eqs 1–5, SI), are summarized in Table .

2. Summary of Calculated Physical, Chemical, and Electronic Properties Derived from the HOMO and LUMO Energies of the Proposed Azobenzene Compounds.

cmpd	ionization potential (IP)	electron affinity (EA)	chemical hardness (η)	chemical potential (μ)	electrophilicity index (ω)	electronegativity (χ)	dipole moment (Debye)
trans-1	6.09	2.56	1.76	–4.33	5.31	4.33	6.19
cis-1	5.93	2.27	1.83	–4.10	4.59	4.10	8.78
trans-2	5.62	2.59	1.52	–4.10	5.55	4.10	6.91
cis-2	5.59	2.38	1.61	–3.98	4.94	3.98	9.24
trans-3	6.09	2.54	1.78	–4.32	5.24	4.32	6.25
cis-3	5.92	2.25	1.84	–4.08	4.54	4.08	6.10
trans-Rh-3	5.43	2.59	1.42	–4.01	5.66	4.01	5.02
cis-Rh-3	5.40	2.36	1.52	–3.88	4.96	3.88	6.41
trans-5	5.74	2.60	1.57	–4.17	5.53	4.17	9.19
cis-5	5.68	2.35	1.66	–4.02	4.85	4.02	8.72

As anticipated, compounds 1 and 3 exhibit the highest η values (1.86 eV), approximated using Koopmans’ theorem, which is consistent with their large HOMO–LUMO gaps and indicative of lower chemical reactivity. However, the limitations in the utilizing η to predict experimental ionization potentials should be noted. Conversely, compound 5 displays the smallest η (1.39 eV) and the highest μ values, suggesting an enhanced propensity for chemical interactions. Importantly, compound 5 also shows the highest combined χ and ω values, indicating a greater tendency for electron uptake and stronger interaction potential with biomolecular targets such as hemozoin. All compounds exhibited IP and EA values within the typical ranges observed for FDA-approved drugs (IP: 4–15 eV; EA: −3 to 7 eV), suggesting an overall favorable pharmacological profile.

The impact of introducing a photoisomerizable azobenzene moiety into a known hemozoin inhibitor scaffold was then assessed using molecular docking. To evaluate binding interactions with hemozoin, the molecular docking simulations were performed using the AutoDock software package, employing a 3 × 3 × 3 ‘supercell’ model of hemozoin to mimic the biomolecular receptor environment. For each ligand, the lowest energy binding pose was retained for analysis, shown in Figures and S2, and the binding affinities detailed in Table .

3.

Docking simulations of (A) trans-1, (B) cis-1, (C) trans-2, and (D) cis-2 with the hemozoin crystal, showing the lowest energy binding conformation.

3. Binding Affinities of the Investigated Compounds to the Hemozoin Crystal.

azobenzene	affinity (kcal/mol)	ΔE (kcal/mol)	ΔE (kJ/mol)
trans-1	–11.6	2.2	9.2
cis-1	–9.4
trans-3	–11.0	2.0	8.4
cis-3	–9.0
trans-2	–10.5	2.8	11.7
cis-2	–7.7
trans-5	–12.4	3.3	13.8
cis-5	–9.1

The docking results revealed consistent trends, with the trans isomers demonstrating stronger binding affinities (<−10.5 kcal/mol) than their cis counterparts, which is attributed to their larger surface area and planar conformation, favoring π-interactions with the porphyrin core of hemozoin. These results parallel established findings for quinoline-based antimalarials, , wherein π-interactions are critical for β-hematin inhibition. ,, Furthermore, the in silico binding affinities correlate with the global reactivity descriptors, where, notably, compound 5, identified as the most chemically reactive via DFT analysis (Table ), also displayed the strongest binding to the hemozoin crystal. This synergy between electron structure and binding propensity supports the hypothesis that conformational changes (e.g., cis–trans isomerization) can interfere with a compound’s ability to interact favorably with a target, directly influencing its pharmacological efficacy. However, since the trans form is both the more stable isomer and the one that binds most strongly to hemozoin, these results highlight important challenges for the current design strategy and provide valuable insights that can guide the optimization for future photopharmacological agents. Nevertheless, given that photoisomerization modulates these binding interactions, subsequent experimental validation was undertaken to assess the light-dependent inhibition of β-hematin formation to confirm the computational predictions.

Synthesis

Based on the results obtained from the DFT studies and the insights provided by molecular docking, we proceeded with the synthesis of the proposed compounds and prepared similar derivatives to allow for a more in-depth comparison. Compound 1 was previously reported, although an alternative synthetic route was followed to obtain this compound and its isosteric ferrocenyl derivative (2) (Scheme S1). Given that a mixture of constitutional isomers (regioisomers) are possible upon cyclometalation of 1 (Figure ), a propyl-containing derivative of 1 (3) was designed, with the overall synthetic route shown in Scheme .

4.

Possible constitutional isomers upon cyclometalation of compound 1.

1. Synthesis of the Propyl-Containing 2-Phenyl and 2-Ferrocenylbenzimidazole Azobenzene Compounds, 3 and 4, and the Corresponding Platinum Group Metal (PGM) Cyclometalated Complexes (Rh-3, Ru-3, and Ir-3) .

^a Reagents and conditions: (i) Propylamine (2.0 equiv), K₂CO₃ (1.5 equiv), Et₃N (1.5 equiv), ACN, 40 °C, 21 h; (ii) NH₄Cl (10 equiv), Zn (20 equiv), MeOH, 25 °C, 84 h; (iii) benzaldehyde (1.2 equiv), NaHSO₃ (5.0 equiv), EtOH:H₂O (1:1), 76 °C, 22 h 30 min or ferrocene carboxaldehyde (1.2 equiv), MgSO₄ (6.0 equiv), TFA (0.1 equiv), MeOH, 65 °C, 13 h; (iv) nitrosobenzene (1.5 equiv), AcOH (1.0 mL), CHCl₃, 25 °C, 23–41 h; (v) [Ru­(p-cymene)­Cl₂]₂ or [Rh­(Cp*)­Cl₂]₂ or [Ir­(Cp*)­Cl₂]₂ (0.5 equiv), NaOAc (2.1 equiv), dry DCM, r.t., 64 h.

The alkylation reaction involved reacting 1-bromo-2,4-dinitrobenzene and propylamine via a nucleophilic aromatic substitution (S_NAr) (Scheme ). Subsequent reduction of the nitro group afforded the triamine compound, which was used without further purification or drying for the subsequent cyclocondensation reaction with either benzaldehyde or ferrocenecarboxaldehyde, to afford the phenyl or ferrocenyl precursors, respectively. Finally, the formation of the azobenzene involved reacting the benzimidazole precursor with nitrosobenzene in the presence of AcOH, to form the final benzimidazole-based azobenzene compounds, 3 and 4. The corresponding cyclometalated complexes (Rh-3, Ru-3, and Ir-3) were prepared by reacting (E)-2-phenyl-6-(phenyldiazenyl)-1-propyl-1H-benzo­[d]­imidazole (3) with the various metal dimers, via C–H activation, in the presence of sodium acetate, to afford the respective complexes in moderate yields (60–76%).

The synthesis of the quinoline-azobenzene compounds (5 and 6) involved reacting 4,7-dichloroquinoline with either 4-hydroxyazobenzene or 4-aminoazobenzene (Scheme ). Compound 5 precipitated from the solution in its protonated form (5H). To obtain the neutral compound, the collected solid was dissolved in MeOH, and NaOAc added to induce precipitation of 5 (neutral).

2. Synthesis of the Quinoline-Azobenzene Compounds, 5H and 6 .

^a Reagents and conditions: (i) 4-Aminoazobenzene (1.0 equiv), EtOH, 78 °C, 16 h; (ii) 4-hydroxyazobenzene (1.2 equiv), K₂CO₃ (1.2 equiv), DMF, 150 °C, 21 h.

The synthesized compounds were fully characterized using ¹H and ¹³C­{¹H} NMR spectroscopy (Figures S3–S18), high-resolution mass spectrometry (HR-MS) (Figures S19–S30), and the purity of the compounds ascertained using either liquid chromatography–mass spectrometry (LC–MS) or high-performance liquid chromatography (HPLC) analysis (Figures S31–S35). Furthermore, the molecular structures of selected compounds were confirmed in the solid state using single-crystal X-ray diffraction.

Suitable crystals of the azobenzene-containing quinoline compounds, 5 and 6, were obtained by the slow evaporation of a saturated methanolic or dichloromethane solution, respectively (Figure ), with the molecular structures confirming the isolation of the trans-isomers at room temperature. This was expected due to the increased thermodynamic stability of the trans-product relative to the metastable cis-isomer, which is further supported by the relative energies obtained and reported in Table S1. For compound 6, the substituted and unsubstituted phenyl rings of the azobenzene moiety are not coplanar with the azo (NN) bond, with torsion angles of −14.7° (C_12A-C_13A-N_2A-N_3A) and −19.0° (N_2A-N_3A-C_16A-C_21A). The trans-azobenzene moiety within this compound is thus not planar. Additionally, the phenoxy group was found to have a torsion angle of −78.1° relative to the quinoline scaffold (C₉–O₁-C₁₀-C_15A), suggesting the azobenzene group is twisted nearly perpendicular relative to the quinoline moiety. On the other hand, in compound 5, the substituted and unsubstituted phenyl rings are nearly coplanar with the NN bond, with torsion angles of 3.0° (C₁₂–C₁₃-N₃–N₄) and −1.2° (N₃–N₄-C₁₆–C₂₁). The azobenzene group is also twisted relative to the quinoline moiety, although to a lesser extent than observed in 6, by nearly 45°. The increased planarity observed for this compound, compared to 6, corroborates the docking studies discussed earlier (vida supra) and the subsequent β-hematin results (vida infra), with a more planar compound shown to interact more strongly with the heme crystal layers, due to increased π-interactions, resulting in the increased inhibition of β-hematin formation. Crystals of the Rh­(III) cyclometalated complex, Rh-3, were obtained from the slow evaporation of a saturated chloroform solution. Complex Rh-3 was shown to crystallize in the triclinic space group P, and from the molecular structure (Figure ) is shown to adopt the well-documented “three-legged piano stool” configuration. , Further crystallographic data and refinement parameters for compounds 5, 6, and Rh-3, are summarized in Tables S2–S5 along with selected bond lengths and angles.

5.

Molecular structure of the quinoline-azobenzene compounds, 5 and 6, and the Rh­(III) cyclometalated complex, Rh-3, with solvent molecules and hydrogen atoms omitted for clarity. Ellipsoids are shown at 50% probability level.

Photophysical Studies

The synthesized azobenzene compounds were analyzed using UV–vis spectrometry to investigate their absorption profiles and photoswitching behavior, both of which are highly dependent on the molecular architecture of the system. − To optimize trans → cis photoisomerization, the key experimental parameters, namely the irradiation wavelength and exposure time, were systematically determined to achieve maximal conversion. The choice of irradiation wavelength is critical for successful photoswitching and is a primary consideration in developing azobenzene-based compounds for photopharmacology. Additionally, thermal relaxation studies provided insights into the half-life of the metastable cis isomer and photoswitching cycles used to examine the photorecyclability.

The UV–vis spectra of the thermodynamically favored trans isomers display intense absorption maxima (λ_max) between 335 and 405 nm, corresponding to symmetry-allowed π → π* transitions, and a weaker, longer-wavelength band assigned to symmetry-forbidden n → π* transitions, which is typical for azobenzene derivatives. − Irradiation with light of the appropriate (optimal) wavelength (Table S6) resulted in a significant reduction in the π → π* absorbance band, accompanied by a concomitant increase in the n → π* absorption, characteristic of trans→cis isomerization. The wavelength of light that induced the greatest absorption change at λ_max was selected as the optimal wavelength for future experiments with that given compound (Table S6). In all cases, the spectral changes observed after irradiation were consistent with those of azobenzene derivatives, indicating efficient trans–cis isomerization. Furthermore, to ensure maximal conversion to the cis isomer, the ideal length of irradiation was determined by irradiating the compounds with the determined optimal wavelength of light, for varying lengths of time, with the results summarized in Table S6.

Modifying an azobenzene to allow for absorption at a longer wavelength often results in a concomitant increase in the cis → trans interconversion, due to a lower thermal barrier, resulting in faster thermal relaxation rates. Aminoazobenzene derivatives such as 5 are generally classified by their relatively rapid cis → trans isomerization. Aminoazobenzenes exhibit significantly higher rates of thermal back isomerization as the electron-donating substituents reduce the thermal isomerization barrier by increasing the electron density in the π*-orbital. ,

Since the goal was to investigate the ability of the synthesized azobenzenes to inhibit β-hematin formation in their trans and cis forms (before and after irradiation, respectively), the time required for complete thermal relaxation was examined (Figures b and S39), since the presence (or absence) of the cis isomer is likely to influence biological activity. Thermal relaxation refers to the spontaneous thermal isomerization of the metastable cis-isomer to the thermodynamically stable trans-isomer. Maintaining the sample at 37 °C allowed for a more accurate determination of the thermal relaxation rate at the incubation temperature, providing insight into whether the metastable form persists throughout the assay. A longer thermal relaxation time generally indicates higher isomerization barriers. However, at higher temperatures, the isomerization barrier tends to decrease, which accelerates the rate of thermal isomerization relaxation. , For all synthesized compounds, except for the ferrocenyl-based complexes (vida infra), the spectrum obtained after light irradiation gradually reverts to the initial spectrum until the absorbance at the initial λ_max is restored. The benzimidazole-based compounds (1 and 3) and PGM-containing complexes (Rh-3, Ru-3, and Ir-3) exhibit thermal relaxation times exceeding 5 h, suggesting the cis product is present throughout the β-hematin assay. In contrast, the quinoline-based compound (5) undergoes relatively fast thermal relaxation, with nearly immediate conversion to the trans isomer after irradiation (Table S6).

6.

(A) UV–Vis spectrum of compound 1, in DMSO, at thermal equilibrium (red line) and the photostationary state after irradiation with 380 nm light (blue line), (B) Changes in the UV spectra during thermal relaxation of 1 at 37 °C over 13 h after irradiation in DMSO, (C) ¹H NMR spectra of 1 in DMSO (12.5 mg/mL solution) before irradiation (bottom) and after irradiation (top) for 2 min with light at 380 nm. The blue denotes the signals attributed to the Z-isomer. (D) Switching cycles of 1, in DMSO, showing the λ_max absorbance following alternating irradiations with 380 and 530 nm light, demonstrating reversible photoisomerization cycles, with minimal photofatigue resistance.

These findings support the in silico studies through which the cis forms of the benzimidazole-based compounds (1 and 3) were found to be lower in energy compared to 5 (Table S1), suggesting increased reactivity or reduced thermal stability of the quinoline-based ligand. The trans ΔE _HL of compounds 1 and 3 (3.53 and 3.55 eV, respectively) versus 5 (3.15 eV) further aligns with the experimentally observed bathochromic shift of the absorption maximum in 5 (Table S6). Additionally, the larger HOMO–LUMO gap for the cis forms of compounds 1 and 3 vs 5 (Table ) also supports the shorter thermal relaxation rate observed for compound 5.

Interestingly, the ferrocene-containing compounds (2 and 4) did not thermally relax after irradiation (Figure S39). Over time, the n → π* absorption band of the ferrocenyl complex 4 decreased in intensity, as had occurred for the other compounds. However, the absorption band observed at 292 nm in the spectrum of the trans form was shifted after irradiation and did not return to the initial wavelength during the investigated thermal relaxation period. Furthermore, irradiation with longer wavelength light sources also did not result in photoisomerization back to the trans form. Speculating that the ferrocenyl entity may be responsible for this inconsistency, we conducted further investigations using 2-ferrocenylbenzimidazole, which lacks the photoswitchable azo functionality. It was initially expected that irradiation would not affect the UV spectrum of the compound due to the absence of the photoswitchable (azo) functionality. However, this was not the case, as the UV spectra showed noticeable changes when light sources below 650 nm were used (Figure S40). To further investigate these changes, NMR spectra of the compound were recorded before and after irradiation (Figure S41). In the ¹H NMR spectrum of 2-ferrocenylbenzimidazole prior to irradiation, the signals were clearly resolved. However, following exposure to 530 nm light, the signals notably broadened, with progressive broadening as irradiation time increased. Notably, the signal intensity of the ferrocenyl protons decreased with an additional broadened signal observed at approximately 6.5 ppm and a shoulder at approximately 3 ppm. Based on a study by Morrison et al., this additional signal is likely attributed to “free” cyclopentadiene as a result of photoinstability. It has previously been reported that ferrocenyl ketones can undergo photoaquation, in wet DMSO, upon exposure to a light source with λ > 280 nm, to produce a monocyclopentadienyliron cation, the free anionic ligand, and free cyclopentadiene. Although these ferrocenyl compounds lack the carbonyl functionality, it is possible that there are aspects that may influence their photostability due to the presence of the Fe²⁺ metal center. It is also known that the presence of a paramagnetic metal ion induces line broadening, which in this case, may have been triggered by light irradiation.

Moreover, all the compounds, bar the quinoline-based compounds, which displayed relatively fast thermal relaxation rates, and the ferrocenyl compounds, which do not revert to the initial form following irradiation, demonstrated excellent photofatigue resistance/photoswitchable stability, with no significant changes in the absorbance values of λ_max observed following alternating irradiation cycles, with the photoswitching graphs shown in Figures d and S42.

A comparative analysis of the photophysical properties and the previously computed HOMO–LUMO energy gaps revealed notable correlations (Tables and ). Specifically, there are relationships between the ΔE _HL and the wavelength used to induce trans to cis isomerization, as well as the rate of thermal relaxation. A smaller HOMO–LUMO energy gap facilitates an electronic excitation from the HOMO (n-orbital) to the LUMO (π*-antibonding orbital) using lower energy, longer-wavelength light. Isomerization then most likely occurs from this S₁ excited state through rotation, which has been shown to be barrierless. , Conversely, larger energy gaps necessitate higher-energy, shorter-wavelength irradiation to achieve isomerization. This trend is exemplified by the benzimidazole-based compounds 1 and 3, for which the trans isomers required relatively high-energy light (λ = 380 nm) for isomerization, consistent with their relatively large HOMO–LUMO gaps of 3.53 and 3.55 eV, respectively. These findings align with theoretical expectations, as larger HOMO–LUMO gaps correlate with increased stability, demanding greater energy input to perturb the electronic configuration. In contrast, compounds with smaller HOMO–LUMO gaps, such as compound 5, were photoswitched using lower-energy, longer-wavelength light, consistent with their higher chemical reactivity and lower excitation thresholds. Furthermore, a correlation was observed between the HOMO–LUMO gap of the cis isomers and their thermal relaxation rates, in which smaller gaps were associated with faster relaxation to the thermodynamically favored trans form (e.g., cis-5 exhibited the fastest relaxation kinetics and had the smallest ΔE _HL).

4. Selected Photophysical Properties of the Representative Compounds.

	photophysical studies
cmpd	optimal λ light (nm) (trans→cis)	thermal relaxation
1	380	≈13 h
2	530	not applicable
3	380	≈15 h
Rh-3	380	≈44 h
5	405	240 s

An exception to these trends was noted for the rhodium complex (Rh-3). Despite its trans isomer exhibiting the smallest HOMO–LUMO gap, it required high-energy irradiation (λ = 380 nm) for isomerization. Additionally, the cis-Rh complex, while having a smaller HOMO–LUMO gap compared to the other compounds, demonstrated slow thermal relaxation, with complete thermal relaxation occurring after approximately 44 h. This discrepancy may arise from metal–ligand interactions and the involvement of d-orbitals, which alter the electronic transition dynamics beyond simple n → π* considerations.

In Vitro Erythrocytic Antiplasmodial Activity

Before studying the impact of the photoinduced conformational change on the β-hematin inhibitory activity, it was essential to determine if the compounds possess antiplasmodial efficacy, a prerequisite for investigations into hemozoin inhibition. The synthesized azobenzene compounds were evaluated for their in vitro blood stage antiplasmodial activity against the drug-sensitive PfNF54 and multidrug resistant PfK1 strains using a Plasmodium lactate dehydrogenase (pLDH) assay, with the IC₅₀ values and corresponding resistance indices (RI) listed in Table . The aminoquinoline-based compound (5) is the most potent among the tested compounds against the NF54 strain, with an IC₅₀ value of 1.52 μM, while ligand 1 displays the most potent activity of the tested organic compounds against the K1 strain (IC₅₀ = 4.83 μM). Notably, the oxygen-containing derivative, 6, is at least four times less active than its amine-containing counterpart 5, in both strains. This diminished activity was not surprising as it is well reported that the secondary amine, at the 4-position of the quinoline scaffold, is crucial for antiplasmodial activity. Upon analysis of the benzimidazole-based compounds, the phenyl derivatives display comparable activity to their ferrocenyl counterparts, (1 vs 2 and 3 vs 4), with the incorporation of the ferrocenyl group shown to have no notable impact on the activity against the drug-sensitive strain. However, the substitution of the phenyl ring for the ferrocenyl moiety resulted in a slight reduction in the antiplasmodial activity in the K1 strain. Interestingly, the docking studies also show that the binding affinities of 1 and 2 differ, suggesting that the docking model used may be more relevant to understanding the behavior of the K1 strain than the drug-sensitive strain. Introducing the propyl chain diminished the activity slightly, against both strains, for both the phenyl and ferrocenyl derivatives (1 vs 3 and 2 vs 4), also correlating with the docking results. Furthermore, in both the NF54 and K1 strains, the rhodium-containing complex was notably less active than the uncoordinated compound (3) and its ruthenium and iridium congeners. Incorporation of the ruthenium and iridium PGMs resulted in these complexes (Ru-3 and Ir-3) displaying low micromolar potency, up to nearly four times better than the free ligand (3) against the sensitive strain. Finally, all the complexes display less cross-resistance between the drug-sensitive and -resistant strains than CQ (Resistance index, RI = 10.95). Generally, the benzimidazole-based compounds, 1, 2, 3 and 4, display RI values less than or equal to one, suggesting increased or equipotent activity against the K1 strain and thus limited or no cross-resistance.

5. In Vitro Antiplasmodial Activity of the Azobenzene-Based Compounds against the Drug-Sensitive PfNF54 and Drug-Resistant PfK1 Strains.

cmpd	PfNF54 IC50 ± SEM (μM)	PfK1 IC50 ± SEM (μM)	RI
1	6.29 ± 0.09	4.83 ± 0.24	0.77
2	6.64 ± 0.53	6.68 ± 0.21	1.01
3	9.96 ± 0.74	6.64 ± 0.29	0.67
4	11.04 ± 1.14	10.00 ± 0.33	0.91
Ru-3	3.38 ± 0.40	4.41 ± 0.23	1.30
Ir-3	2.57 ± 0.33	4.79 ± 0.25	1.86
Rh-3	17.13 ± 1.16	21.37 ± 1.23	1.25
5	1.52 ± 0.17	6.49 ± 0.32	4.27
6	47.77 ± 4.00	26.50 ± 1.01	0.55
CQ	0.022 ± 0.004	0.241 ± 0.005	10.95

^a(IC₅₀ K1/IC₅₀ NF54).

Inhibition of β-Hematin Formation (Cell-Free Study)

The target for many quinoline- ,,− and benzimidazole-based , compounds is proposed to be hemozoin inhibition and the ability of a potential drug candidate to inhibit hemozoin formation can be measured using the cell-free NP40-mediated β-hematin formation inhibition assay, which mimics the conditions in the acidic DV of the parasite. ,, Consequently, the synthesized azobenzene-containing compounds (1–6 and Ru-3, Rh-3, and Ir-3) were investigated for their ability to inhibit β-hematin (synthetic hemozoin) formation as a potential mechanism of antiplasmodial action. The compounds were analyzed without prior irradiation, i.e., in which the trans form was dominant, and after irradiation with light of the appropriate wavelength to yield the cis form, with the IC₅₀ values of the compounds shown in Table .

6. IC₅₀ Values Obtained from the β-Hematin Inhibition Studies of the Azobenzene-Containing Compounds and CQDP, before and after Light Irradiation.

	IC50 ± SEM (μM)
cmpd	before irradiation	after irradiation
1	35.90 ± 1.32	68.05 ± 2.53 ,
2	21.85 ± 1.18	17.69 ± 1.29
3	142.7 ± 7.83	183.32 ± 7.62 ,
4	35.80 ± 1.90	31.74 ± 3.49
Ru-3	>500	>500
Ir-3	>500	>500
Rh-3	>500	>500
5	15.78 ± 1.07	18.92 ± 1.38
6	84.68 ± 6.65	93.81 ± 6.60
2-Phenylbenz	>500	>500
2-Ferrocenylbenz	49.37 ± 0.21	18.57 ± 2.59
CQ	14.34 ± 0.76	17.70 ± 1.27
		16.18 ± 0.76
		14.72 ± 1.98

^aIrradiation wavelength = 380 nm.

^bIrradiation wavelength = 405 nm.

^cIrradiation wavelength = 530 nm.

^d p < 0.001.

^e p < 0.05 (Statistical significance between the irradiated and nonirradiated activity determined using a Welch’s two sample t test).

Generally, the compounds display moderate β-hematin formation inhibitory activity, with compound 5 found to possess activity comparable to the known hemozoin formation inhibitor drug, CQ (Table ). Compound 1 displayed moderate activity (IC₅₀ = 35.90 μM) before irradiation, with the incorporation of the propyl chain shown to significantly hinder the activity, with 3 having an IC₅₀ value nearly four times greater than 1. A similar trend was observed for the ferrocenyl analogue (2 vs 4).

Most notably, irradiation of 1 and 3 with 380 nm light reduced their inhibitory activity, with 1, for example, showing a nearly 2-fold reduction in activity following irradiation (IC₅₀ = 35.90 vs 68.05 μM). The trans isomer of 1 is relatively planar, and it is speculated that irradiation of this compound to the resulting cis-isomer, which presents a more angular geometry, is less able to “slide” between the heme layers and inhibit β-hematin formation. Hemozoin contains dimers consisting of heme linked by reciprocating monodentate carboxylate linkages. These cross-linked dimers are held together by π–π interactions, forming a network of sheets stabilized by iron-carboxylate linkages. Known β-hematin inhibitors, such as CQ, form a π–π complex with heme, ,, coprecipitating with the heme under acidic conditions to form a covalent CQ-heme complex, with the formation of this complex ultimately preventing the biocrystallization to hemozoin. Therefore, as shown by the molecular docking studies (vide supra), the isomerization of compounds 1 and 3 to the cis form reduces its ability to form a favorable π–π stabilized complex with heme and effectively inhibit the formation of β-hematin.

The amine-containing quinoline ligand (5) was over five times more potent than the hydroxyl-based ligand (6) before light irradiation. Notably, the IC₅₀ value of compound 5 after irradiation is almost within the standard error of the mean (SEM) of the IC₅₀ value obtained before irradiation. These results do not demonstrate that the two forms are equally active; rather, as the earlier thermal relaxation studies revealed, there is almost immediate thermal relaxation following irradiation. Thus, it is likely that very early in the assay, all the cis product had thermally relaxed to the thermodynamically stable trans form, which explains the similar activity observed before and after irradiation.

The PGM-containing compounds (Ru-3, Ir-3, and Rh-3) were the least active of the synthesized compounds and were shown to be inactive at the highest tested concentration, with IC₅₀ values >500 μM, agreeing with structurally similar complexes reported in the literature. Despite these complexes showing antiplasmodial activity (Table ), their inability to inhibit β-hematin formation suggests that the compounds do not act via this pathway but rather impart their activity via an alternative mechanism of action. The inability of these compounds to act via this mechanism may be due to the bulky 3D nature of the metal fragment, again resulting in the compounds being unable to “slide” between the layers and effectively prevent crystal formation.

CQ, used as the control in the experiment, was also studied in the absence and presence of light irradiation. Since CQ lacks photoswitchable groups in its chemical structure, its activity was expected to remain relatively unchanged between irradiated and nonirradiated conditions. Despite slight discrepancies in the IC₅₀ values obtained during the light experiments, all values fall within the SEMs and only slightly outside the SEM of the IC₅₀ value obtained without light, confirming that light does not affect CQ’s inhibitory activity. A further control experiment was conducted using 2-phenylbenzimidazole and 2-ferrocenylbenzimidazole, which lack the azo functionality, to assess their β-hematin inhibitory activity under both light and dark conditions. Similar to CQ, no changes in the activity were expected, as the absence of a photoswitchable group should prevent structural changes upon light irradiation. However, there is a notable difference in the activity of 2-ferrocenylbenzimidazole before and after irradiation, as expected based on the previously discussed NMR results. The enhanced activity is thus likely due to the changes induced by light irradiation (photoinstability), attributed to the presence of the ferrocenyl moiety. , Finally, since the entire plate is irradiated after serial dilution of the compound, but before the addition of the hemin solution, the effect of light on the NP40:DMSO:H₂O solution, alone, revealed that the structural integrity of the NP40 is retained following irradiation, with no significant changes in the UV spectra (Figure S43) observed before and after irradiation with the various light sources.

These cell-free results collectively suggest that this series of hybrids may act as hemozoin formation inhibitors; however, the conformational changes induced by light irradiation are not favorable for the inhibition of β-hematin formation. The reduced planarity of the cis isomer diminishes its ability to form favorable π–π interactions with heme, as suggested by the initial docking studies (Figure ), which confirmed that the trans isomers are more effective β-hematin inhibitors, displaying higher binding affinities with the crystal face than their cis counterparts. Moreover, molecular docking simulations identified ligand 5 as having the strongest binding affinity with hematin (Figure ), which is consistent with the fact that it is the most potent inhibitor of β-hematin formation (Table ).

Conclusion and Future Work

Quinoline- and benzimidazole-based compounds containing photoisomerizable azobenzene units were designed and investigated using density functional theory (DFT) calculations. These studies confirmed that the trans isomers are thermodynamically more stable than their metastable cis counterparts. Analysis of their electronic properties revealed clear correlations between low chemical hardness, high electronegativity, and high electrophilicity indices. Molecular docking studies demonstrated that the trans isomers exhibited stronger binding affinities to the hemozoin crystal, attributed to their enhanced planarity compared to the torsionally strained cis isomers. Encouraged by these computational results, the title compounds and selected derivatives were synthesized for further study. Photophysical investigations revealed that extending the π-conjugated system via quinoline and benzimidazole substitution induced a bathochromic shift of the absorption maxima relative to the unsubstituted azobenzene. Incorporation of an electron-donating amino group in compound 5 resulted in a further red shift, allowing effective excitation at 405 nm, albeit with an associated increase in thermal relaxation rate, which corresponds with the calculated HOMO–LUMO energy gaps of the cis isomers, whereby compound 5 was shown to have the smallest gap. Conversely, the benzimidazole-based compounds required shorter-wavelength light (380 nm) for isomerization but exhibited slower thermal relaxation. All compounds display moderate antiplasmodial activity, warranting the investigation into their β-hematin inhibitory activity. The activity of the benzimidazole ligands (1 and 3) were notably affected by light irradiation, suggesting the influence of conformational changes on activity. The reduced activity after irradiation is possibly due to the more structurally strained angular compound being unable to effectively interact with heme through π-interactions and form stable interactions. These results reveal that the configuration of a compound plays a crucial role in its β-hematin inhibitory activity, particularly emphasizing the importance of compound planarity when designing a compound with this specific target in mind. It further underscores that light can be used to manipulate the configuration, which in turn can alter (favorably or unfavorably) the interaction of the photoswitchable compound with its intended target and concomitant bioactivity. Future work into the rational design of quinoline- and benzimidazole-azobenzene compounds would involve red-shifting the λ_max of the compounds, so longer wavelength light can be used to induce isomerization. This can be achieved by incorporating electron-donating substituents (e.g., OMe) or push–pull substituents at the para positions. To enhance the pharmacological efficacy, a hybrid approach could be utilized, with the two pharmacophoric groups linked by the azo functionality. This may also result in a more notable structural change upon irradiation and potentially prominent differences in the pharmacological efficacy of the two isomeric forms. For this study in particular, future work will involve cellular heme fractionation studies to determine whether the resulting cell-free β-hematin activity will translate in vitro. Finally, the effects of photoisomerization on the in vitro antiplasmodial activity can also be investigated, as the two isomeric forms may display appreciable differences in activity, although not necessarily exerting their effect via hemozoin inhibition.

Experimental Procedures

Computational Studies and Molecular Docking

The geometries of the azobenzene compounds were optimized in vacuo using density functional theory (DFT). Geometry optimizations were carried out with the B3LYP functional and the def-SVP basis set. Single-point energy calculations were subsequently performed using the def2-TZVPP basis set for all atoms. All calculations were performed with the Gaussian16 software package, and the optimized structures were visualized using Chemcraft. Molecular docking simulations were performed to support the experimental findings of the β-hematin inhibition assay. The simulations were conducted using a 3 × 3 × 3 ‘supercell’ of hemozoin as the biomolecular receptor, utilizing the AutoDock Vina software package. As required by AutoDock Vina, all receptor and ligand structures were converted to the pdbqt format, an extension of pdb that includes partial charges (Q) and atom types (T). The search space was defined to enclose the entire crystal surface, with a center at coordinates (x, y, z): (13.5, 22.5, 12) and dimensions of (x, y, z): (40, 40, 40). Each ligand was docked with an exhaustiveness level of 8, and only the lowest energy binding mode was retained. All docking results were analyzed and visualized using PyMOL.

Chemicals and General Methods

All reagents and solvents were purchased from commercial sources (Sigma-Aldrich, Merck, Combi-Blocks and KIMIX) and were used without further purification. All solvents were of analytical grade, with some dried over molecular sieves. All reactions were carried out under an inert argon atmosphere using standard Schlenk line techniques unless otherwise stated. Reactions were monitored by thin-layer chromatography (TLC) using aluminum-backed Merck precoated silica-gel 60 F254 plates and compounds visualized by ultraviolet light at 254 nm. Column chromatography was conducted using 60 Å silica gel (70–230 mesh). The synthesis of 5-nitro-2-phenyl-1H-benzo­[d]­imidazole, 2-phenyl-1H-benzo­[d]­imidazole-5-amine, and (E)-2-phenyl-5-(phenyldiazenyl)-1H-benzo­[d]-imidazole (1) have been previously reported, although alternative or modified synthetic procedures were used to obtain the desired compounds.

Nuclear magnetic resonance (NMR) spectra were recorded on a Bruker XR600 MHz spectrometer (¹H at 599.95 MHz and ¹³C­{¹H} at 151.0 MHz), a Bruker Topspin GmbH (¹H at 400.22 MHz and ¹³C­{¹H} at 100.65 MHz), or Bruker 300 MHz (¹H at 300.08 MHz) spectrometer. These were equipped with a Bruker Biospin GmbH casing and sample injector at 30 °C and tetramethylsilane (TMS) was used as the internal standard. Coupling constants are reported in Hz and chemical shifts are reported in ppm relative to residual solvent signals. Infrared (IR) spectroscopy was performed on a PerkinElmer Spectrum 100 FT-IR spectrometer using Attenuated Total Reflectance (ATR) with vibrations measured in units of cm^–1. Electronic absorption spectra were recorded on an Agilent Cary 8454 Photodiode Array UV/Visible spectrometer. LC–MS analysis was performed using an Agilent 1260 Infinity Binary Pump, an Agilent 1260 Infinity diode array detector (DAD), an Agilent 1290 Infinity Column Compartment, an Agilent 1260 Infinity Standard Autosampler, and an Agilent 6120 Quadrupole (Single) MS system, with an APCI/ESI multimode ionization source. The purities were determined by an Agilent LCMS/MS system using an X-bridge C18 5 μm column (4.6 × 150 mm); organic phase: 10 mM ammonium acetate (pH 3.7) in HPLC-grade methanol, aqueous phase: 10 mM ammonium acetate (pH 3.7) in HPLC-grade water; flow rate = 1.20 mL/min; detector: PDA.

The purity of the complexes were determined using an analytical Agilent HPLC 1260 infinity II, equipped with an Agilent DAD 1260 UV/vis detector and an Agilent Pursuit 5 C18 column (5.0 μM, 150 mm × 4.6 mm). The compounds were eluted using a mixture of solvent A (0.1% TFA in H2O) and solvent B (MeOH) at a flow rate of 0.5 mL/min. The gradient elution conditions were as follows: 10% solvent B between 0 and 2 min, 10–90% solvent B between 2 and 8 min, 90% solvent B between 8 and 9 min. Purity was determined at 254 nm. All compounds were confirmed to have >97% purity.

X-ray Crystallography

Single-crystal X-ray diffraction data were collected on a Bruker D8 Venture diffractometer using graphite-monochromated Mo–Kα radiation (χ = 0.71073 Å). Data collection was carried out at 100(2) K for 6, at room temperature for 5, and at 173(2) K for Rh-3. The temperature was controlled by an Oxford Cryostream cooling system (Oxford Cryostat). Cell refinement and data reduction were performed using the program SAINT. The data were scaled and absorption correction performed using SADABS. The structure was solved by direct methods using SHELXS-97 and refined by full-matrix least-squares methods based on F ² using SHELXL-2014 and using the graphics interface program X-Seed. , The programs X-Seed and POV-Ray were used to prepare molecular graphic images.

Compound 5: All non-hydrogen atoms were refined anisotropically. All hydrogen atoms, except H2 and H1G, were placed in idealized positions and refined in riding models with U_iso assigned 1.2 or 1.5 times U_eq of their parent atoms and the C–H bond distances were constrained in the range 0.93–0.96 Å. The hydrogen atoms H2 and H1G were located in the difference density maps and refined independently. The structure was refined to R factor of 0.0874.

Compound 6: All non-hydrogen atoms were refined anisotropically. All hydrogen atoms were placed in idealized positions and refined in riding models with U_iso assigned 1.2 times U_eq of their parent atoms and the C–H bond distances were constrained to 0.95 Å. Part of the molecule was disordered over two positions, with refined partial occupancy factors of 0.621(5) for part A and 0.379(5) for part B. The structure was refined to R factor of 0.0367.

Complex Rh-3: All non-hydrogen atoms were refined anisotropically. All hydrogen atoms were placed in idealized positions and refined in riding models with U_iso assigned 1.2 or 1.5 times U_eq of their parent atoms. The C–H bond distances were constrained to 0.95, 0.98, and 0.99 Å, respectively. The structure was refined to R factor of 0.0340.

Synthesis

Known precursors, 5-nitro-2-phenyl-1H-benzo­[d]­imidazole and 2-phenyl-1H-benzo­[d]­imidazole-5-amine, were used in the synthesis of (E)-2-phenyl-5-(phenyldiazenyl)-1H-benzo­[d]­imidazole (1). The ferrocenyl-based precursor, 5-nitro-2-ferrocenyl-1H-benzo­[d]­imidazole, was used in the synthesis of 2. These known compounds were synthesized following alternative synthetic procedures (Supporting Information Scheme S1).

(E)-2-Ferrocenyl-5-(phenyldiazenyl)-1H-benzo­[d]­imidazole (2)

2-Ferrocenyl-1H-benzo­[d]­imidazole-5-amine (0.156 g, 0.490 mmol) was dissolved in a mixture of chloroform (8.75 mL) and acetic acid (12.5 mL). Nitrosobenzene (0.0578 g, 0.539 mmol) was then added, and the solution stirred at r.t. for 4 days and 16 h. The solvent of the clear red-brown solution was removed by creating an azeotrope with EtOH. DCM was added to the remaining residue and the insoluble material filtered. The solvent of the filtrate was then removed under reduced pressure and purified via column chromatography, using EtOAc as the eluent. The appropriate fraction was collected and purified again using DCM:MeOH as the eluent. This yielded 2 as a bright orange powder. Yield: 76% (0.152 g; 0.373 mmol); R _f (DCM:MeOH, 95:5) = 0.45; LC-MS (m/z): t _R = 0.93 min, Experimental = 407.1 (100% purity, [M + H]⁺, calc. = 407.1); ¹H NMR (600 MHz, [D₄]-MeOD): δ­(ppm) = 8.09 (br s, 1H, H_b), 7.92 (d, 3H, H_a,g, ³ J _H–H = 7.7 Hz), 7.64 (d, 1H, H_c, ³ J _H–H = 8.5 Hz), 7.55 (t, 2H, H_h, ³ J _H–H = 7.4 Hz), 7.52–7.46 (m, 1H, H_i); ¹³C­{¹H} NMR (100 MHz, [D₄]-MeOD): δ (ppm) = 158.30, 154.16, 150.11, 131.72, 130.24, 123.61, 119.43, 73.32, 71.87, 70.92, 68.83; HR-MS (ESI (+), m/z): Found = 407.0973 ([M + H]⁺), Calculated = 407.0953.

2,4-Dinitro-N-propylamine

Propylamine (1.72 mL, 20.9 mmol), K₂CO₃ (2.20 g, 15.9 mmol) and Et₃N (2.18 mL, 15.7 mmol) were added to a stirring solution of 1-bromo-2,4-dinitrobenzene (2.58 g, 10.4 mmol) in ACN (30.0 mL). The reaction mixture was then heated to 40 °C and stirred for 21 h and 30 min. The mixture was diluted with H₂O (50.0 mL), the aqueous solution extracted with EtOAc (3 × 100.0 mL), and organic layers dried over anhydrous Na₂SO₄. Following filtration, the solvent was removed to afford the desired product as a bright yellow solid. Yield: 100% (2.36 g; 10.4 mmol); R_f (DCM) = 0.83; ¹H NMR (300 MHz, [D₄]-MeOD): δ (ppm) = 9.02 (s, 1H, H_a), 8.27 (d, 1H, H_b, ³ J _H–H = 9.6 Hz), 7.16 (d, 1H, H_c, ³ J _H–H = 9.6 Hz), 3.45 (t, 2H, H_d, ³ J _H–H = 7.1 Hz), 1.77 (sextet, 2H, H_e), 1.06 (t, 3H, H_f, ³ J _H–H = 7.3 Hz); ¹³C­{¹H} NMR (100 MHz, [D₄]-MeOD): δ­(ppm) = 149.8, 136.8, 131.4, 131.0, 124.8, 115.8, 45.9, 23.0, 11.6; FT-IR (ATR): (v _max/cm^–1) = 3365 (N–H), 1619 (CN), 1582 (NO_2 asym.), 1334 (NO_2 sym.).

N ¹-Propylbenzene-1,2,4-triamine

NH₄Cl (3.20 g, 59.7 mmol) was added to a stirring solution of 2,4-dinitro- N -propylamine (1.23 g, 5.45 mmol) in anhydrous MeOH (40.0 mL) and the yellow mixture stirred at r.t., under Ar, for 40 min. Zn powder (7.31 g, 112 mmol) was then added and the resulting gray mixture stirred vigorously at r.t., for approximately 84 h (Scheme ). Thereafter, the reaction mixture was filtered through Celite, to remove the zinc, and the MeOH removed under reduced pressure. This yielded the desired product as a dark purple residue. ¹H NMR (300 MHz, [D₄]-MeOD): δ­(ppm) = 6.52 (d, 1H, H_c, ³ J _H–H = 7.8 Hz), 6.27 (s, 1H, H_a), 6.19 (d, 1H, H_b, ³ J _H–H = 7.7 Hz), 2.94–2.78 (m, 2H, H_d), 1.69–1.48 (m, 2H, H_e), 0.94 (t, 3H, H_f, ³ J _H–H = 7.4 Hz).

2-Phenyl-1-propyl-1H-benzo­[d]­imidazol-5-amine

Benzaldehyde (0.500 mL, 4.88 mmol) and NaHSO₃ (2.22 g, 21.3 mmol) were dissolved in a mixture of EtOH:H₂O (1:1, 20.0 mL), which was stirred at 76 °C for 50 min. After cooling, N ¹ -propylbenzene-1,2,4-triamine (1.02 g, 4.07 mmol) in minimal MeOH was added to the reaction mixture, which was heated to 76 °C and stirred for a further 22 h 30 min. Thereafter, the solvent was reduced and the remaining aqueous residue extracted with DCM (3 × 75.0 mL). The organic fractions were combined, dried over anhydrous Na₂SO₄, filtered, and the solvent of the filtrate removed to afford a dark red residue. The crude product was purified via column chromatography using Et₂O:EtOAc as the eluent, to yield the desired product as a red viscous residue. Yield: 24% (0.243 g; 0.968 mmol); R _f (Et₂O) = 0.11, (Et₂O:EtOAc, 1:1) = 0.36; LC–MS (m/z): t _R = 0.60 min, Experimental = 252.2 (100% purity, [M + H]⁺, calc. = 252.2); ¹H NMR (300 MHz, [D₁]-CDCl₃): δ­(ppm) = 7.72–7.64 (m, 2H, H_c), 7.56–7.45 (m, 3H, H_a,b), 7.20 (d, 1H, H_e, ³ J _H–H = 8.6 Hz), 7.11 (s, 1H, H_d), 6.74 (d, 1H, H_f, ³ J _H–H = 8.3 Hz), 4.13 (t, 2H, H_j, ³ J _H–H = 7.6 Hz), 3.67 (br s, 2H, NH₂), 1.90–1.76 (m, 2H, H_k), 0.85 (t, 3H, H_l, ³ J _H–H = 7.2 Hz); ¹³C­{¹H} NMR (100 MHz, [D₁]-CDCl₃): δ (ppm) = 153.83, 144.31, 142.27, 130.97, 129.87, 129.60, 129.38, 128.75, 112.94, 110.59, 104.91, 46.47, 23.33, 11.36; FT-IR (ATR): (v _max/cm^–1) = 3330, 3211 (N–H), 1626 (CC and CN).

2-Ferrocenyl-1-propyl-1H-benzo­[d]­imidazol-5-amine

N ¹ -Propylbenzene-1,2,4-triamine (0.563 g, 3.41 mmol), in minimal MeOH (5.00 mL), ferrocenecarboxaldehyde (0.880 g, 4.11 mmol), and a catalytic amount of TFA were dissolved in MeOH (25.0 mL). Anhydrous MgSO₄ (2.44 g, 20.2 mmol) was then added and the dark red reaction mixture refluxed for 13 h. The mixture was filtered by gravity, to remove the MgSO₄, and the solvent of the filtrate removed to give a dark red residue. To this crude residue, 1 M NaOH (30.0 mL) was added and the mixture extracted using DCM (3 × 50.0 mL). The organic fractions were combined, washed with brine (50.0 mL), followed by H₂O (50.0 mL), and dried over Na₂SO₄. Following filtration, the solvent of the filtrate was removed to afford the crude product which was purified via column chromatography, using EtOAc:Et₂O as the eluent. Compound 4 was isolated as a red-brown powder. Yield: 28% (0.345 g; 0.959 mmol); LC-MS (m/z): t _R = 0.71 min, Experimental = 360.1 (100% purity, [M + H]⁺, calc. = 360.1); ¹H NMR (300 MHz, [D₁]-CDCl₃): δ (ppm) = 7.11 (d, 1H, H_b, ³ J _H–H = 8.3 Hz), 7.05 (s, 1H, H_a), 6.68 (d, 1H, H_c, ³ J _H–H = 8.4 Hz), 4.87 (s, 2H, H_g), 4.44 (s, 2H, H_h), 4.30 (t, 2H, H_d, ³ J _H–H = 7.7 Hz), 4.18 (s, 5H, H_i), 3.27 (br s, 2H, NH), 2.02–1.83 (m, 2H, H_e), 1.08 (t, 3H, H_f, ³ J _H–H = 7.3 Hz); ¹³C­{¹H} NMR (100 MHz, [D₁]-CDCl₃): δ (ppm) = 152.77, 144.28, 142.07, 130.29. 112.05, 109.56, 104.22, 74.38, 69.95, 69.77, 68.98, 46.03, 23.74, 11.64; HR-MS (ESI (+), m/z): Found = 360.1169 ([M + H]⁺), Calculated = 360.1157.

General Method for Compounds 3 and 4

Nitrosobenzene (1.5 equiv) and AcOH (2.5 equiv) were added to stirring solution of 2-phenyl-1-propyl-1 H -benzo­[ d ]­imidazol-5-amine or 2-ferrocenyl-1-propyl-1 H -benzo­[ d ]­imidazol-5-amine (1.0 equiv) in CHCl₃ (10.0 mL). The yellow-brown solution was stirred at r.t. for either 21 h (3) or 42 h (4). The solvent was removed and the red crude product purified via column chromatography, using 100% EtOAc as the eluent. The first fraction contained the desired product and unreacted nitrosobenzene. The collected fraction was purified via column chromatography once again, using 100% DCM, with the second fraction (spot) containing the desired product.

(E)-2-Phenyl-5-(phenyldiazenyl)-1-propyl-1H-benzo­[d]­imidazole (3)

Nitrosobenzene (0.0998 g, 0.932 mmol) and AcOH (0.142 mL, 2.48 mmol) were added to stirring solution of 2-phenyl-1-propyl-1 H -benzo­[ d ]­imidazol-5-amine (0.156 g, 0.621 mmol) in CHCl₃ (10.0 mL). The isolated fraction was dried in vacuo to give 3 as an orange solid. Yield: 83% (0.175 g; 0.513 mmol); R _f (EtOAc) = 0.92, R _f (DCM) = 0.41; LC-MS (m/z): t _R = 1.14 min, Experimental = 341.2 (96% purity, [M + H]⁺, calc. = 341.2); ¹H NMR (600 MHz, [D₁]-CDCl₃): δ (ppm) = 8.43 (s, 1H, H_d), 8.04–7.91 (m, 3H, H_f,g), 7.79–7.67 (m, 2H, H_c), 7.61–7.39 (m, 7H, H_a,b,h,i,e), 4.23 (t, 2H, H_j, ³ J _H–H = 7.3 Hz), 1.97–1.77 (m, 2H, H_k), 0.88 (t, 3H, H_l, ³ J _H–H = 7.3 Hz); ¹³C­{¹H} NMR (100 MHz, [D₁]-CDCl₃): δ (ppm) = 155.55, 152.86, 149.12, 143.46, 137.92, 130.59, 130.43, 130.10, 129.42, 129.17, 128.93, 122.83, 117.52, 116.69, 110.39, 46.67, 23.38, 11.35; FT-IR (ATR): (v _max/cm^–1) = 1604 (CC and CN).

(E)-2-Ferrocenyl-5-(phenyldiazenyl)-1-propyl-1H-benzo­[d]­imidazole (4)

Nitrosobenzene (0.113 g, 1.05 mmol) and AcOH (0.50 mL) were added to a stirring solution of 2-ferrocenyl-1-propyl-1 H -benzo­[ d ]­imidazol-5-amine (0.244 g, 0.678 mmol) in CHCl₃ (10.0 mL). The isolated fraction was dried in vacuo to give 4 as a dark red solid. Yield: 70% (0.214 g; 0.478 mmol); R _f (DCM) = 0.14; LC–MS (m/z): t _R = 1.07 min, Experimental = 449.1 (100% purity, [M + H]⁺, calc. = 449.1); ¹H NMR (300 MHz, [D₁]-CDCl₃): δ (ppm) = 8.40 (s, 1H, H_a), 7.95 (d, 3H, H_b,j, ³ J _H–H = 7.5 Hz), 7.53 (t, 2H, H_k, ³ J _H–H = 7.0 Hz), 7.47 (d, 1H, H_m, ³ J _H–H = 6.5 Hz), 7.40 (d, 1H, H_c, ³ J _H–H = 8.7 Hz), 4.97 (s, 2H, H_g), 4.51 (s, 2H, H_h), 4.39 (t, 2H, H_d, ³ J _H–H = 7.6 Hz), 4.22 (s, 5H, H_i), 2.07–1.88 (m, 2H, H_e), 1.11 (t, 3H, H_f, ³ J _H–H = 7.2 Hz); ¹³C­{¹H} NMR (100 MHz, [D₁]-CDCl₃): δ (ppm) = 155.13, 152.86, 149.14, 142.82, 138.02, 130.52, 129.14, 122.83, 116.70, 115.69, 109.42, 73.02, 70.54, 70.03, 69.39, 46.32, 23.75, 11.58; FT-IR (ATR): (v _max/cm^–1) = 1710, 1606 (CC and CN); HR-MS (ESI (+), m/z): Found = 449.1439 ([M + H]⁺), Calculated = 449.1423.

General Method for Cyclometalated Complexes Ru-3, Ir-3, and Rh-3

Sodium acetate (2.0 equiv) was added to a stirring solution of 3 (1.0 equiv) in dry DCM (10.0 mL) and the solution stirred for approximately 15 min. Thereafter, the appropriate dimer (0.5 equiv) was added and the solution stirred at 25 °C, under argon, for between 64 and 76 h. The resulting solution was filtered through Celite and the solvent reduced to a minimal (ca. 1.00 mL). Excess Et₂O was then added, resulting in the precipitation of the desired complex which was collected by filtration and washed with copious amounts of cold Et₂O, followed by hexane.

2-Phenylbenzimidazole Azobenzene Cyclometalated Ru­(II) p-Cymene Complex (Ru-3)

Sodium acetate (0.0355 g, 0.433 mmol), 3 (0.0711 g, 0.209 mmol), and [Ru­(p-cymeme)­Cl₂]₂ (0.0639 g, 0.104 mmol). Excess Et₂O was added and the solution stored at −2.0 °C for 48 h. Following filtration and washing, the desired product was isolated as a dark brown crystalline solid. Yield: 39% (24.8 mg; 0.041 mmol); ¹H NMR (300 MHz, [D₁]-CDCl₃): δ (ppm) = 8.53 (s, 1H, H_g), 8.37 (d, 1H, H_a, ³ J _H–H = 7.4 Hz), 8.00 (d, 3H, H_f,h, ³ J _H–H = 7.9 Hz), 7.65 (d, 1H, H_d, ³ J _H–H = 7.6 Hz), 7.61–7.49 (m, 3H, H_i,j), 7.46 (d, 1H, H_e, ³ J _H–H = 8.9 Hz), 7.29–7.17 (m, 1H, H_b), 7.09 (t, 1H, H_c, ³ J _H–H = 7.2 Hz), 5.94 (d, 1H, H_q’, ³ J _H–H = 5.5 Hz), 5.75 (d, 1H, H_q, ³ J _H–H = 5.4 Hz), 5.45 (d, 1H, H_r’, ³ J _H–H = 5.5 Hz), 5.18 (d, 1H, H_r, ³ J _H–H = 5.3 Hz), 4.55–4.42 (m, 1H, H_k), 4.42–4.28 (m, 1H, H_k’), 2.42–2.25 (m, 1H, H_p), 2.13 (s, 3H, H_s), 2.08–1.91 (m, 2H, H_m), 1.03 (t, 3H, H_n, ³ J _H–H = 7.3 Hz), 0.95 (d, 3H, H_o, ³ J _H–H = 6.7 Hz), 0.80 (d, 3H, H_o’, ³ J _H–H = 6.7 Hz); ¹³C­{¹H} NMR (100 MHz, [D₁]-CDCl₃): δ (ppm) = 184.70, 159.58, 152.85, 149.35, 141.83, 140.87, 138.24, 133.66, 130.89, 129.28, 129.23, 124.35, 122.95, 122.77, 118.77, 112.58, 110.37, 101.44, 99.33, 89.60, 89.25, 82.42, 81.44, 46.62, 31.04, 23.19, 22.71, 21.94, 19.14, 11.38; FT-IR (ATR): (v _max/cm^–1) = 1607 (CC), 1580 (CN); HR-MS (ESI (+), m/z): Found = 575.1758 ([M-Cl]⁺), Calculated = 575.1749.

2-Phenylbenzimidazole Azobenzene Cyclometalated Ir­(III)-Cp* Complex (Ir-3)

Sodium acetate (0.0505 g, 0.616 mmol), 3 (0.0945 g, 0.278 mmol), and [Ir­(Cp*)­Cl₂]₂ (0.111 g, 0.139 mmol). The addition of excess Et₂O resulted in the immediate precipitation of the desired product, as a dark red powder. Yield: 90% (0.0877 g; 0.125 mmol); ¹H NMR (300 MHz, [D₁]-CDCl₃): δ­(ppm) = 8.30 (s, 1H, H_g), 8.10–7.90 (m, 4H, H_a,f,h), 7.68 (d, 1H, H_d, ³ J _H–H = 7.7 Hz), 7.62–7.42 (m, 4H, H_i,j,e), 7.30–7.17 (m, 1H, H_b), 7.09 (t, 1H, H_c, ³ J _H–H = 7.7 Hz), 4.61–4.34 (m, 2H, H_k), 2.19–1.98 (m, 2H, H_m), 1.81 (s, 15H, H_o), 1.10 (t, 3H, H_n, ³ J _H–H = 7.2 Hz); ¹³C­{¹H} NMR (100 MHz, [D₁]-CDCl₃): δ (ppm) = 166.44, 163.77, 152.72, 149.44, 140.11, 138.25, 137.28, 133.87, 131.01, 130.87, 129.23, 124.09, 122.96, 122.19, 121.04, 110.32, 109.85, 88.39, 46.67, 23.34, 11.50, 9.90; FT-IR (ATR): (v _max/cm^–1) = 1604 (CC), 1583 (CN); HR-MS (ESI (+), m/z): Found = 667.2413 ([M-Cl]⁺), Calculated = 667.2413.

2-Phenylbenzimidazole Azobenzene Cyclometalated Rh­(III)-Cp* complex (Rh-3)

Sodium acetate (0.0323 g, 0.394 mmol), 3 (0.0670 g, 0.197 mmol), and [Rh­(Cp*)­Cl₂]₂ (0.0608 g, 0.098 mmol). The addition of excess Et₂O resulted in minimal precipitation. The solution was stored at 3.4 °C for approximately three hours to promote further precipitation and the desired product was obtained as an orange powder. Yield: 76% (45.7 mg; 0.0745 mmol); ¹H NMR (300 MHz, [D₁]-CDCl₃): δ (ppm) = 8.37 (s, 1H, H_g), 8.03 (d, 1H, H_a, ³ J _H–H = 7.6 Hz), 7.99–7.90 (m, 3H, H_f,h), 7.68–7.40 (m, 5H, H_d,i,j,e), 7.35–7.22 (m, 1H, H_b), 7.11 (t, 1H, H_c), 4.43 (t, 2H, H_k, ³ J _H–H = 7.0 Hz), 2.16–1.98 (m, 2H, H_m), 1.72 (s, 15H, H_o), 1.09 (t, 3H, H_n, ³ J _H–H = 7.0 Hz); ¹³C­{¹H} NMR (100 MHz, [D₁]-CDCl₃): δ­(ppm) = 181.59 (d, J _Rh–C = 31.4 Hz), 159.60, 152.81, 149.28, 140.55, 138.48, 138.07, 134.45, 130.83, 130.14, 129.23, 123.86, 122.96, 120.41, 110.98, 110.35, 95.98, 94.23, 46.80, 23.27, 11.47, 9.98; FT-IR (ATR): (v _max/cm^–1) = 1604 (CC), 1579 (CN); HR-MS (ESI (+), m/z): Found = 577.1831 ([M-Cl]⁺), Calculated = 577.1838.

(E)-7-Chloro-N-(4-(phenyldiazenyl)­phenyl)­quinolin-4-amine (5)

A mixture of 4,7-dichloroquinoline (0.251 g, 1.27 mmol) and 4-aminoazobenzene (0.250 g, 1.27 mmol) were dissolved in EtOH (20.0 mL) and the solution heated under reflux for 16 h and 40 min. An evident orange precipitate was observed. The solvent was then reduced to a minimal, the precipitate collected by suction filtration, and washed with cold EtOH. This yielded the protonated product as a bright orange powder. To obtain the deprotonated form, the solid was stirred in MeOH, with excess NaOAc, and the desired bright orange precipitate collected by suction filtration. Yield: 84% (0.382 g; 1.06 mmol); R_f (DCM:MeOH, 90:10) = 0.76; LC-MS (m/z): t _R = 0.83 min, Experimental = 359.1 (96% purity, [M + H]⁺, calc. = 359.1); ¹H NMR (300 MHz, [D₆]-DMSO): δ­(ppm) = 9.48 (br s, 1H, NH), 8.61 (d, 1H, H_a, ³ J _H–H = 4.0 Hz), 8.43 (d, 1H, H_d, ³ J _H–H = 8.8 Hz), 8.02–7.90 (m, 3H, H_c,g), 7.87 (d, 2H, H_h, ³ J _H–H = 7.4 Hz), 7.69–7.48 (m, 6H, H_b,f,i,j), 7.29 (d, 1H, H_e, ³ J _H–H = 3.9 Hz); ¹³C­{¹H} NMR (100 MHz, [D₆]-DMSO): δ (ppm) = 152.17, 149,74, 147.35, 146.36, 144.27, 134.27, 131.00, 129.48, 127.84, 125.55, 124.75, 124.33, 122.37, 120.54, 119.20, 104.63.

(E)-7-Chloro-4-(4-(phenyldiazenyl)­phenoxy)­quinoline (6)

K₂CO₃ (0.369 g, 2.67 mmol) was added to a stirring solution of 4-hydroxyazobenzene (0.489 g, 2.47 mmol) in anhydrous DMF (4.00 mL). Thereafter, 4,7-dichloroquinoline (0.407 g, 2.06 mmol) was added and the reaction mixture heated under reflux (150 °C) for 21 h. The solvent was then removed under reduced pressure by creating an azeotrope with ethanol. Water (40.0 mL) was added to the remaining residue and the aqueous solution extracted with DCM (70.0 mL). The organic layer was washed with brine (40.0 mL) and dried over anhydrous Na₂SO₄. Following filtration, the solvent was removed and the remaining red residue purified via column chromatography, using DCM as the eluent, yielding the desired product as a bright orange-red crystalline solid. Yield: 28% (0.210 g; 0.584 mmol); R _f (DCM) = 0.21; LC-MS (m/z): t _R = 1.37 min, Experimental = 360.1 (100% purity, [M + H]⁺, calc. = 360.1); ¹H NMR (300 MHz, [D₁]-CDCl₃): δ­(ppm) = 8.73 (d, 1H, H_a, ³ J _H–H = 3.1 Hz), 8.30 (d, 1H, H_d, ³ J _H–H = 8.8 Hz), 8.15 (br s, 1H, H_c), 8.07 (d, 2H, H_g, ³ J _H–H = 7.7 Hz), 7.97 (d, 2H, H_h, ³ J _H–H = 6.8 Hz), 7.64–7.44 (m, 4H, H_b,i,j), 7.33 (d, 2H, H_f, ³ J _H–H = 7.7 Hz), 6.67 (d, 1H, H_e, ³ J _H–H = 4.3 Hz); ¹³C­{¹H} NMR (100 MHz, [D₆]-DMSO): δ (ppm) = 160.33, 156.33, 153.09, 151.91, 149.79, 149.41, 135.16, 131.69, 129.56, 127.58, 127.29, 124.93, 123.81, 122.62, 121.59, 119.52, 106.07; FT-IR (ATR): (v _max/cm^–1) = 1614 (CN).

In Vitro Asexual Antiplasmodial Activity

Test samples were screened in technical triplicate, for three or four biological replicates, for their in vitro blood stage antiplasmodial activity against the drug-sensitive (NF54) and multidrug-resistant (K1) strains of P. falciparum. Continuous in vitro cultures of asexual erythrocyte stages of P. falciparum were maintained using a modified method of Trager and Jensen. A quantitative assessment of the in vitro antiplasmodial activity was determined via the parasite lactate dehydrogenase (pLDH) assay, using a modified method described by Makler and Hinrichs. The test samples were prepared to a 20 mg/mL stock solution in 100% DMSO. Stock solutions were stored at −20 °C. Further dilutions were prepared in complete medium on the day of the experiment. Samples were tested as a suspension if not completely dissolved. Chloroquine diphosphate (CQDP) was used as the reference drug in all experiments. Test samples were tested at a starting concentration of either 100, 10, or 1 μg/mL, which was then serially diluted 2-fold, in complete medium, to give 10 concentrations. The same dilution technique was used for all samples. A full dose–response was performed for all compounds to determine the concentration inhibiting 50% of parasite growth (IC₅₀ value). The IC₅₀ values were obtained from full dose–response curves, using a nonlinear dose–response curve fitting analysis, via GraphPad Prism v.5 software.

β-Hematin Formation Inhibitory Activity

The β-hematin formation inhibition assay was modified from the method reported by Sandlin et al. , Stock solutions of the respective test compounds were prepared to 10 mM, in 100% DMSO, with the exception of CQ (used in the salt form, CQDP), which was prepared in Milli-Q water. A 100 μL aliquot of a water/NP-40 (305.5 μM)/DMSO solution, in a v/v ratio of 7/2/1, was added to all wells in columns 1–11. Milli-Q water (140 μL), NP-40 (305.5 μM, 40 μL), and a 20 μL aliquot of the test compound were added to the wells in column 12. NP-40 is a detergent used to mediate the formation of β-hematin. The solution in column 12 was then serially diluted 2-fold through to column 2. Column 1 contained 0 μM of the test compound and thus served as a blank. A 25 mM stock solution of hemeatin was then prepared, by dissolving 16.3 mg of hemein in 1.0 mL of DMSO, and sonicated for one minute. A 178.8 μL aliquot of this hemeatin suspension was added to 20.0 mL of a 2 M acetate buffer (pH = 4.8) and a 100 μL of this solution added to each well. The plates were then covered and incubated for 5 h, at 37 °C. The assay was analyzed using the pyridine-ferrochrome method, developed by Egan and co-workers. A solution containing water/acetone/2 M HEPES buffer (pH 7.4)/pyridine, in a v/v ratio of 2/2/1/5, was then prepared and 32 μL of this solution added to each well, followed by 60 μL of acetone. The absorbance values were then measured at 405 nm, the results plotted using GraphPad Prism (v5), and the IC₅₀ values obtained using a sigmoidal dose–response curve fitting analysis.

Glossary

Abbreviations

WHO

World Health Organization

ACT

Artemisinin-based combination therapy

MMV

medicines for malaria venture

Hz

Hemozoin

Hb

Hemoglobin

FQ

ferroquine

CQ

chloroquine

Cp

cyclopentadienyl

DV

digestive vacuole

MDR

multidrug-resistant

ABS

asexual blood stage

SNAr

nucleophilic aromatic substitution

LC-MS

liquid chromatography–mass spectrometry

TFA

trifluoroacetic acid

HR-MS

high resolution-mass spectrometry

HPLC

high-performance liquid chromatography

3D

3-dimensional

PfCRT

Plasmodium falciparum chloroquine resistance transporter

pLDH

Plasmodium lactate dehydrogenase

CQS

chloroquine-sensitive

CQR

chloroquine-resistant

CQDP

chloroquine diphosphate

RI

resistance index

SI

selectivity index

TLC

thin layer chromatography

TMS

tetramethylsilane

IR

infrared

NMR

nuclear magnetic resonance.

The Supporting Information is available free of charge at https://pubs.acs.org/doi/10.1021/acsomega.5c09966.

• The study provides experimental data, synthesis methods, and analysis of the compounds' properties. It includes detailed insights into the synthesis, structural properties, and photochemical behavior of azobenzene-based β-hematin inhibitors, highlighting their potential in malaria treatment and photopharmacology (PDF)

Crystallographic details: CCDC numbers: 2434931 (5), 2434937 (6), 2434942 (Rh-3).

T.M.G.: Investigation, data curation, writingoriginal draft, validation, and formal analysis. T.K.T.: Computational and docking studies, data curation. G.S.S.: Conceptualization, resources, writingreviewing and editing, supervision, project administration, and funding acquisition. S.A.B.: Supervision. C.E.: Writingreviewing and editing, computational and docking studies, data curation, supervision.

We gratefully acknowledge and thank the University of Cape Town and the National Research Foundation of South Africa under a Competitive Programme for Rated Researchers (GSS UID: 129288) for financial support.

The authors declare no competing financial interest.