DNA Intercalating Near-Infrared Luminescent Lanthanide Complexes Containing Dipyrido[3,2-a:2′,3′-c]phenazine (dppz) Ligands: Synthesis, Crystal Structures, Stability, Luminescence Properties and CT-DNA Interaction

In order to create near-infrared (NIR) luminescent lanthanide complexes suitable for DNA-interaction, novel lanthanide dppz complexes with general formula [Ln(NO3)3(dppz)2] (Ln = Nd3+, Er3+ and Yb3+; dppz = dipyrido[3,2-a:2′,3′-c]phenazine) were synthesized, characterized and their luminescence properties were investigated. In addition, analogous compounds with other lanthanide ions (Ln = Ce3+, Pr3+, Sm3+, Eu3+, Tb3+, Dy3+, Ho3+, Tm3+, Lu3+) were prepared. All complexes were characterized by IR spectroscopy and elemental analysis. Single-crystal X-ray diffraction analysis of the complexes (Ln = La3+, Ce3+, Pr3+, Nd3+, Eu3+, Er3+, Yb3+, Lu3+) showed that the lanthanide’s first coordination sphere can be described as a bicapped dodecahedron, made up of two bidentate dppz ligands and three bidentate-coordinating nitrate anions. Efficient energy transfer was observed from the dppz ligand to the lanthanide ion (Nd3+, Er3+ and Yb3+), while relatively high luminescence lifetimes were detected for these complexes. In their excitation spectra, the maximum of the strong broad band is located at around 385 nm and this wavelength was further used for excitation of the chosen complexes. In their emission spectra, the following characteristic NIR emission peaks were observed: for a) Nd3+: 4F3/2 → 4I9/2 (870.8 nm), 4F3/2 → 4I11/2 (1052.7 nm) and 4F3/2 → 4I13/2 (1334.5 nm); b) Er3+: 4I13/2 → 4I15/2 (1529.0 nm) c) Yb3+: 2F5/2 → 2F7/2 (977.6 nm). While its low triplet energy level is ideally suited for efficient sensitization of Nd3+ and Er3+, the dppz ligand is considered not favorable as a sensitizer for most of the visible emitting lanthanide ions, due to its low-lying triplet level, which is too low for the accepting levels of most visible emitting lanthanides. Furthermore, the DNA intercalation ability of the [Nd(NO3)3(dppz)2] complex with calf thymus DNA (CT-DNA) was confirmed using fluorescence spectroscopy.


Introduction
Lanthanide chemistry is a field of inorganic chemistry that is constantly evolving, particularly in recent decades, as a growing number of synthesized compounds are widely used in different areas. One of the most fascinating aspects of trivalent lanthanides is their unique luminescence properties [1][2][3]. They show characteristic narrow line-like emission peaks, ranging from the UV-visible (UV/Vis) to for measurements of analogous structures. A detailed luminescence study of the [Nd(NO 3 ) 3 2 ] complexes was carried out and is presented in this paper. Additionally, the stability of the compounds in solution is verified by UV-Vis spectroscopy. Last, the intercalation ability of the [Nd(NO 3 ) 3 (dppz) 2 ] with calf thymus DNA (CT-DNA) was tested using fluorescence spectroscopy.
The IR spectra of all the complexes showed characteristic bands at the following wavenumber positions: around 3095 cm −1 (C-H stretching vibrations), 1600-1585 and 1500-1400 cm −1 (C-C and C-N stretching vibrations in the ring), 900-675 cm −1 (out-of-plane bends).
An isostructural compound of the Eu 3+ complex has previously been reported, however, this latter compound crystallized in the different space group C2/c, and featured inclusions of an extra dppz ligand and diethylether molecule in the crystal packing [25].

Photoluminescence Study
The [Ln(NO 3 ) 3 (dppz) 2 ] complexes containing spectroscopically active lanthanide ions were tested for their luminescence properties. All measurements were carried out in the solid state. It was observed that only the [Nd(NO 3 ) 3 (dppz) 2 ], [Er(NO 3 ) 3 (dppz) 2 ], [Yb(NO 3 ) 3 (dppz) 2 ] and [Eu(NO 3 ) 3 (dppz) 2 ] complexes displayed luminescence properties when excited into the maximum of the broad band originating from the dppz ligand. The NIR luminescence properties of the Nd 3+ , Er 3+ and Yb 3+ complexes are discussed in detail in this paper. For the sake of completeness, the excitation and emission spectra of the [Eu(NO 3 ) 3 (dppz) 2 ] complex are presented in the SI (Figures S1 and S2). The singlet and triplet energy levels of the dppz ligand have been previously reported in the literature (37000 and 18600 cm −1 ) [22]. It is known that the intersystem crossing process becomes effective when ∆E ( 1 ππ*− 3 ππ*), the energy difference between the first excited singlet and triplet levels of the ligand, is more than 5000 cm −1 [26]. For the dppz ligand, the energy gap between the 1 ππ* and 3 ππ* amounts to 18400 cm −1 , which indicates an effective intersystem crossing process. A study of the ligand's triplet energy level and the resonant energy levels of the various lanthanides allows to predict whether efficient ligand-to-metal energy transfer can occur between the chosen ligand and a particular lanthanide ion. The ligand triplet energy level must be obviously higher than the resonant energy level of the lanthanide to result in adequate ligand to metal energy transfer. Of course, most likely the energy is not transferred directly to the emitting level of the lanthanide, but to higher levels, which then relax to the emitting levels via a nonradiative way. In Scheme 2 we present the singlet and triplet energy levels of dppz and of some selected spectroscopically active lanthanide ions. This clearly shows that the dppz ligand is very well suited for the design of NIR emitting Nd 3+ , Er 3+ and Yb 3+ complexes. On the other hand, the low triplet energy level of dppz makes it rather unfavorable as a sensitizer for most of the visible emitting lanthanide ions (e.g., Sm 3+ , Tb 3+ , Dy 3+ ). Nevertheless, we were able to observe visible red emission for the [Eu(NO 3 ) 3 (dppz) 2 ] complex (studied in solid state, DMF and a DMF-H 2 O mixture; Figures S1 and S2). Scheme 2. Schematic simplified energy level diagram of the dppz ligand and selected lanthanides employed in the study. Favorable energy transfer processes from the dppz to Ln 3+ ions are shown as orange dotted arrows.
Steady-state excitation and emission spectra of the Nd 3+ , Er 3+ and Yb 3+ complexes were recorded at room temperature. For all three complexes, the excitation spectrum consisted of a strong broad band in the region 250-500 nm with a maximum at around 385.0 nm (Figures 2-4). For [Nd(NO 3 ) 3 (dppz) 2 ], additionally several narrow sharp peaks were present, which could be assigned to the f-f transitions of Nd 3+ (Table 1). When exciting the [Nd(NO 3 ) 3 (dppz) 2 ] complex at 385.0 nm the three characteristic NIR emission peaks of Nd 3+ were recorded: 4 F 3/2 → 4 I 9/2 (870.8 nm), 4 F 3/2 → 4 I 11/2 (1052.7 nm), and 4 F 3/2 → 4 I 13/2 (1334.5 nm). The luminescence lifetime of the complex was also recorded. The decay curve of this compound as well as the [Er(NO 3 ) 3 (dppz) 2 ] and [Yb(NO 3 ) 3 (dppz) 2 ] complex could be well fitted using the single exponential equation: where I and I 0 are the luminescence intensities at time t and 0, respectively, and τ is the luminescence lifetime. The decay time of the [Nd(NO 3 ) 3 (dppz) 2 ] complex was determined to be 1624 ns or 1.62 µs ( Figure 5).  . Excitation (left) and emission (right) steady-state spectra of the [Er(NO 3 ) 3 (dppz) 2 ] complex measured at room temperature. The broad emission band is referred to as peak "a" in Table 1. The broad emission band is referred to as peak "a" in Table 1. When exciting the [Er(NO 3 ) 3 (dppz) 2 ] complex at 385.0 nm one strong, broad peak located at 1529.0 nm, which can be assigned to the 4 I 13/2 → 4 I 15/2 "telecom" transition, was observed. The luminescence lifetime was recorded for this complex and was determined as 2295 ns or 2.30 µs ( Figure 6).
Exciting the [Yb(NO 3 ) 3 (dppz) 2 ] complex at 385.0 nm resulted in a strong broad peak with a maximum at 977.6 nm. This peak could be assigned to the 2 F 5/2 → 2 F 7/2 transition of Yb 3+ . The decay time of the complex was calculated to be 9765 nm or 9.77 µs ( Figure 7).   The fact that the decay curves of the three complexes could be well fitted by monoexponential functions indicates the presence of single luminescence sites in the compounds, which is confirmed by the crystal structures.
The 2 ] complexes show good luminescence properties, confirming that the dppz ligand is very well suited for the design of NIR emitting Nd 3+ , Er 3+ and Yb 3+ complexes. The dppz triplet level is well matched with the resonant energy levels of the two lanthanides Nd 3+ and Er 3+ , allowing efficient energy transfer from the ligand to the lanthanide ions. Additionally, there are no water molecules in the first coordination sphere of the lanthanide minimizing non-radiative deactivation. To the best of our knowledge, so far only the luminescence properties of an [Er(bta) 3 (dppz)] complex have been reported for NIR-emitting lanthanide dppz complexes [18]. This compound has a higher decay time compared to our [Er(NO 3 ) 3 (dppz) 2 ] complex. This is quite expected due to the presence of a fluorinated Hbta ligand, as it is known that Er 3+ complexes with fluorinated ligands have exceptionally long decay times (e.g., 16.8 µs for a complex with 1,1,1,3,5,5,5-heptafluoropentane-2,4-dione) [27,28].

UV/Vis Stability Tests
In order to enable the [Ln(NO 3 ) 3 (dppz) 2 ] complexes to have real application in biological environment it is necessary that they remain stable in solution over a certain period of time. Therefore, the stability of the [Nd(NO 3 ) 3 (dppz) 2 ], [Er(NO 3 ) 3 (dppz) 2 ] and [Yb(NO 3 ) 3 (dppz) 2 ] samples was tested over a period of 6 h. As can be found in literature the electronic absorption spectra of the dppz complexes in DMF show a ligand centered π→π* transition at around 265 nm. Two bands at 361 nm and 380 nm assigned to the n→π* transitions of the phenazine moiety are also present in the UV/Vis spectra [22,29]. In order to rule out decomposition of the complexes over time absorption spectral traces of the [Nd(NO 3 ) 3 (dppz) 2 ] in DMF were measured over a period of up to 6 h ( Figure 8). No significant changes are visible in the spectra. Similar experiments were carried out for the [Er(NO 3 ) 3 (dppz) 2 ] and [Yb(NO 3 ) 3 (dppz) 2 ] compounds and exhibited no appreciable changes in the UV/Vis spectra over time ( Figures S3 and S4). Our UV-Vis spectra are recorded in the 280-500 nm region to observe the n→π* transitions of the phenazine moiety. The complexes were also soluble in a mixture of DMF/Tris buffer (5 mM Tris-HCl, 5 mM NaCl, pH 7.2) and weakly soluble in only the Tris buffer ( Figures S4 and S5). To further investigate the stability of these [Ln(NO 3 ) 3 (dppz) 2 ] complexes, we have analyzed the [Eu(NO 3 ) 3 (dppz) 2 ] complex in different environments. It is known, that in europium compounds the splitting of the 5 D 0 → 7 F 0-6 peaks contains information on the europium ion's environment [2]. Therefore, europium is often referred to as a structural probe. We have performed measurements of the [Eu(NO 3 ) 3 (dppz) 2 ] complex in solid, DMF and 50%DMF + 50%H 2 O mixture (see Figures S1 and S2).
All emission spectra have been recorded upon exciting into the maximum of the dppz ligand's absorption band at 370 nm. As can be observed, there are some changes in the emission spectra measured in the different environments, but it is clear that the red emission of the Eu 3+ is retained even in 50%DMF + 50%H 2 O. However, the excitation spectra recorded in the three cases look remarkably similar (apart from the 5 L 6 ← 7 F 0 peak slightly below 400 nm in the solid state spectrum). This clearly indicates that the Eu 3+ ion remains coordinated by at least one of the dppz ligands even after dissolution in DMF and after adding additional distilled water. In order to investigate whether indeed a different complex would be present after dissolution in DMF or DMF/water mixture, the luminescence decay traces were recorded in H 2 O and D 2 O to determine the number of directly coordinated water molecules using the Horrock's equation [30,31] (Figures S7-S10). In fact, DMF/H 2 O 50/50 and DMF/D 2 O 50/50 mixtures were used as the complex isn't sufficiently soluble in pure H 2 O or D 2 O. This yielded a number of coordinated water molecules of 3.73, pointing to the coordination of three or four water molecules to the Eu 3+ ion. Given the fact that at least one dppz ligand should remain coordinated to the Eu 3+ ion, simply replacing one dppz ligand with water molecules would not yield a number of three to four directly coordinated water molecules. As the excitation spectra are almost identical in the three cases, it is more likely that some of the nitrate ions that coordinate to the Eu 3+ ion are replaced by water or DMF molecules (the nitrates remaining in the second coordination sphere for example) upon dissolution, resulting in a local coordination site at the Eu 3+ ion that is less dissimilar to the original one in the solid state. It can therefore be assumed that indeed changes in the complex's composition may occur upon dissolution in DMF or a DMF/water mixture, but most likely this does not involve drastic rearrangements such as the loss of the dppz ligands. As all reported complexes are isostructural, we expect similar behavior of the whole series of compounds.

CT-DNA Interaction Tests
In order to assess if the obtained [Ln(NO 3 ) 3 (dppz) 2 ] complexes show DNA interaction ability we carried out an ethidium bromide (EthB) displacement assay with CT-DNA in the presence of the [Nd(NO 3 ) 3 (dppz) 2 ] complex as an example. EthB is employed as a spectral probe, which shows enhanced emission intensity when intercalated to DNA and reduced emission intensity in the free state in water or buffer medium (due to solvent quenching of the fluorescence). The competitive interaction of the [Nd(NO 3 ) 3 (dppz) 2 ] complex, which could result in the displacement of the CT-DNA intercalated EthB, was monitored using fluorescence spectroscopy.

Characterization
Elemental analysis (C, H, N) was carried out with a Thermo Scientific Flash 2000 Series CHNS/O analyzer.
IR spectra were recorded from 4000 to 650 cm −1 , on a Thermo Scientific FT-IR spectrometer (type Nicolet 6700), equipped with a DRIFTS-cell, using KBr as a non-absorbing matrix.
Photoluminescence measurements were recorded on an Edinburgh Instruments FLSP920 UV/Vis-NIR spectrofluorometer, using a 450 W xenon lamp as the steady state excitation source and a Hamamatsu R5509-72 photomultiplier operating at −80 • C. For the Eu 3+ spectrum, a Hamamatsu R928P PMT was used, for the 200-900 nm range. The time-resolved measurements were performed using a Continuum Surelite I laser (450 mJ @1064 nm), operating at a repetition rate of 10 Hz and using the third harmonic (355 nm) as the excitation source, and the photomultiplier detector mentioned above. For all compounds, the luminescence of solid samples was recorded. Small amounts of the powder were placed between quartz plates. UV/Vis measurements were performed on a Lambda 950 UV/Vis-NIR spectrophotometer from Perkin Elmer employing quartz cuvettes with a 10 mm path length.
NMR spectra were recorded in DMSO-d6, on a Bruker Avance 300 MHz or 500 MHz and all chemical shifts are given relative to tetramethylsilane (TMS).

Synthesis of the Ligand
3.3.1. Synthesis of the Ligand Precursor 1,10-phenanthroline-5,6-dione (Phendione) 1,10-Phenanthroline (55.5 mmol) was added in small portions under stirring to sulfuric acid (65%, 80 mL) and left to dissolve at room temperature. Potassium bromate (62 mmol) was added in portions over a period of 3 h and the mixture was stirred at room temperature for 24 h. Then, the mixture was poured over ice and was neutralized to pH 7 using sodium carbonate. The mixture was extracted with chloroform and the organic phase was dried over magnesium sulfate during the night. The solution was then filtered and evaporated to dryness. The crude product was recrystallized from methanol to obtain the desired yellow product.

Synthesis of the Complexes
The starting lanthanide nitrate salt (0.06 mmol) was dissolved in a mixture of nitromethane/ methanol (1/1, 12 mL) and heated at 40 • C. To this, a solution of dppz ligand (0.12 mmol) in warm methanol (8 mL) was added dropwise with stirring, and the reaction mixture was left for 2 h, at 40 • C. During that period, the desired complex precipitated. The product was filtered off, washed with nitromethane (2 mL), water (1 mL) and ether (2 mL) and dried under reduced pressure. The filtrate was set up for crystallization via vapor diffusion of acetonitrile or isopropanol as antisolvents.

Single Crystal X-ray Diffraction Analysis
For all reported structures, X-ray intensity data were collected, at 100 K, on a Rigaku Oxford Diffraction Supernova Dual Source (Cu at zero) diffractometer equipped with an Atlas charge-coupled device (CCD) detector using ω scans and CuKα (λ = 1.54184 Å) radiation. The images were interpreted and integrated with the program CrysAlisPro (Rigaku Oxford Diffraction) [34]. Using Olex2 [35], the structures were solved by direct methods using the ShelXS structure solution program and refined by full-matrix least-squares on F 2 using the ShelXL program package [36,37]. Nonhydrogen atoms were anisotropically refined and the hydrogen atoms in the riding mode and isotropic temperature factors fixed at 1.2 times U(eq) of the parent atoms. For all reported structures, a correction for diffuse effects, due to the inclusion of disordered water molecules, was made using the SQUEEZE routine in PLATON [38].
Crystal data for compound: C 36  DMF (14 mM) and the appropriate amount was used for the DNA binding study. The competitive binding assay from EthB displacement was carried out by measuring the emission intensities of EthB bound to CT-DNA solution with the gradual increase of the complex concentrations (27.9-406.9 µM; taking into account the dilution). The EthB was excited at 546 nm and observed at 603 nm (emission maximum). EthB showed very weak emission in the Tris-buffer due to fluorescent quenching of the free EthB in buffer. CT-DNA-bound EthB showed significant enhancement in emission intensity.

Conclusions
Herein, we report the synthetic procedures and the spectroscopic and single crystal X-ray diffraction characteristics for a series of lanthanide complexes containing dipyrido[3,2-a:2 ,3 -c]phenazine (dppz) as ligand. Our synthetic approach is novel and is able to afford single crystals suitable for X-ray diffraction analysis for most of the compounds in the lanthanide series. X-ray analysis of the complexes reveals that all complexes are isomorphic, with the lanthanide ion coordinated by two dppz ligands and three nitrate anions. As it could be expected based on the T 1 state of the ligand (18400 cm −1 ) and the resonant energy level of the studied lanthanides (Nd 3+ , Er 3+ , Yb 3+ ), it was shown that the dppz ligand is an appropriate choice for the design of NIR emitting complexes. We have additionally studied the stability of the [Nd(NO 3 ) 3 (dppz) 2 ], [Er(NO 3 ) 3 (dppz) 2 ] and [Yb(NO 3 ) 3 (dppz) 2 ] complexes in solution over a period of time. No visible changes in the UV/Vis spectra were observed up to 6 h. Last, we have confirmed that the exemplary [Nd(NO 3 ) 3 (dppz) 2 ] complex shows DNA intercalation ability through an ethidium bromide displacement assay carried out with CT-DNA. These results point out that such carefully developed lanthanide polypyridyl complexes could serve as potential NIR luminescent probes for bioimaging applications and diagnostics.