Insights from ¹²⁵Te and ⁵⁷Fe nuclear resonance vibrational spectroscopy: a [4Fe–4Te] cluster from two points of view

Abstract

Iron–sulfur clusters are common building blocks for electron transport and active sites of metalloproteins. Their comprehensive investigation is crucial for understanding these enzymes, which play important roles in modern biomimetic catalysis and biotechnology applications. We address this issue by utilizing (Et₄N)₃[Fe₄Te₄(SPh)₄], a tellurium modified version of a conventional reduced [4Fe–4S]⁺ cluster, and performed both ⁵⁷Fe- and ¹²⁵Te-NRVS to reveal its characteristic vibrational features. Our analysis exposed major differences in the resulting ⁵⁷Fe spectrum profile as compared to that of the respective [4Fe–4S] cluster, and between the ⁵⁷Fe and ¹²⁵Te profiles. DFT calculations are applied to rationalize structural, electronic, vibrational, and redox-dependent properties of the [4Fe–4Te]⁺ core. We herein highlight the potential of sulfur/tellurium exchange as a method to isolate the iron-only motion in enzymatic systems.

---

Introduction

Tellurium is a component of many materials of current scientific and technological interest, including thermoelectrics such as PbTe-based materials,^1,2 phase change materials for data storage such as Ge₂Sb₂Te₅,³ and superconductors such as Fe_1+yTe_1−xSe_x and CdTe solar cells.^4,5 Although Te is not an essential trace element in biological systems, certain Te compounds can be bioaccumulated and/or metabolized and some of them have shown or have been proposed to have valuable antibiotic or anticancer applications.⁶ This contrasts with the lighter chalcogens, S and Se, where S is ubiquitous in most living systems, and Se plays an essential role as selenocysteine.^7–9 Only in the late 1980's, storage and transport of semimetal tellurium was observed in biological systems.¹⁰ When the tellurium-tolerant fungi Penicillium chrysogenum was fed with sodium tellurite containing media, Te incorporation was observed to afford telluro-cysteine and -methionine.¹⁰

[4Fe–4S] clusters are one of the most common bioinorganic prosthetic groups, playing crucial roles in many biological processes, such as electron transfer,^11,12 O₂ and NO-sensing, or even as active sites in radical SAM enzymes.^13–17 The role of sulfur atoms in a cofactor is commonly not specifically elucidated, since sulfur spectroscopy of enzymes remains a challenging task due to the sulfur-rich environment of the peptide backbone. Current studies on [4Fe–4S] cluster compounds (or Fe–S interactions in general) are limited to probing the iron sites. Notably, the naturally low abundance of tellurium within an organism makes this element a unique tool to investigate specific processes by following the trace of tellurium, e.g., substituting sulfur(s) at specified locations in biological samples and evaluating the metal–sulfur interaction(s) using difference spectra.

Nuclear resonance vibrational spectroscopy (NRVS),¹⁸ also known as nuclear inelastic scattering (NIS), is a relatively new synchrotron-based technique that yields vibrational spectra from nuclear transitions of an appropriate Mössbauer isotope. The key feature of this technique is that it is only sensitive to the motion of the selected isotope along the direction of the incident X-ray beam. For example, in the case of ⁵⁷Fe-enriched proteins, this allows one to select exclusively the normal modes involving the ⁵⁷Fe nuclei in a background of thousands of other vibrations involving the protein matrix. ⁵⁷Fe-NRVS has already been used to investigate a wide variety of Fe-containing proteins, including rubredoxins, ferredoxins,^19,20 heme proteins,²¹ hydrogenases,^22,23 dioxygenases,²⁴ NO sensors, and nitrogenase.²⁵

In 2010, NRVS was first applied to ¹²⁵Te with its nuclear resonance at 35.49 keV.²⁶ Thanks to the development of sapphire backscattering monochromators,²⁷ spectra with 1.1 meV (<9 cm⁻¹) resolution have been obtained for elemental Te,²⁷ Bi₂Te₃ and Sb₂Te₃,^28–30 as well as GeTe, SnTe, and PbTe.³¹¹²⁵Te presents a unique opportunity for bioinorganic NRVS studies thanks to the feasibility of Te for S substitution. Te can replace bridging S in Fe–S clusters, and it can also replace S to yield tellurocysteine and telluromethionine.

Despite growing popularity for materials science, there have been no reports of ¹²⁵Te-NRVS of biologically relevant complexes to date. In this study we investigate the feasibility of bioinorganic ¹²⁵Te-NRVS studies using tellurium modification of a well-known [4Fe–4S] cluster, resulting in a ⁵⁷Fe and ¹²⁵Te enriched (Et₄N)₃[Fe₄Te₄(SPh)₄] model compound. Dual isotopic labelling allows for deeper insights into the dynamics of the cluster; since certain modes can be strong in one spectrum and weak or absent in the partner spectrum. Here, the unique ability to combine NRVS measurements of both ⁵⁷Fe and ¹²⁵Te in the same sample offers added assessment by density functional theory (DFT) which was applied to analyze the vibrational modes of the [4⁵⁷Fe–4¹²⁵Te] core in a greater detail.

Herein, we provide a benchmark study to reveal the potential of ¹²⁵Te-NRVS and a combination of isotopic labelling of both iron and tellurium as a probe in electron transport processes in future in vivo studies within enzymes' Fe–S systems, such as [FeFe]-hydrogenases and CO-dehydrogenases.

Results and discussion

Synthesis

(Et₄N)₃[⁵⁷Fe₄¹²⁵Te₄(SPh)₄] = 1 was synthesized combining the synthetic procedures reported by Midollini³² and Haase,³³via the reaction of FeCl₂, LiSPh and Li₂Te to obtain black needle-shaped crystals (Scheme 1).

Scheme 1 Schematic representation of the synthesis and ORTEP molecular plot of (Et₄N)₃[Fe₄Te₄(SPh)₄]. The probability level is 50%. Counter ions are omitted for clarity.

The black crystals were analyzed by ¹H NMR spectroscopy and X-ray diffractometry. The proton NMR spectrum (Fig. S1†) shows three paramagnetic signals at 13.5 ppm (meta), −3.1 ppm (ortho) and −4.1 ppm (para) in a 2 : 2 : 1 ratio and are assigned to the aromatic protons of the anionic cluster. In addition, two diamagnetic signals at 3.04 ppm and 1.35 ppm in a ratio of 2 : 3 are observed which can be assigned to the tetraethyl ammonium counter ions. The overall spectrum closely resembles the spectrum presented by Midollini et al.³²

The obtained crystals were of decent quality (R = 2.6) to perform single crystal X-ray analysis, and the resulting structure confirms the successful synthesis of [Fe₄Te₄(SPh)₄]³⁻. Compound 1 crystallizes in the space group Fdd2 (Scheme 1, Table S1†). Since the structural features of the cubic cluster are already thoroughly discussed, we herein refer to the literature.³²

NRVS

Fig. 1a presents the ¹²⁵Te and ⁵⁷Fe partial vibrational density of states (PVDOS) spectra obtained from our NRVS experiments on 1. The ⁵⁷Fe-PVDOS spectrum shows four main features: a small peak at ∼130 cm⁻¹, a global maximum at ∼190 cm⁻¹, a doublet at ∼240–250 cm⁻¹, and weaker features at ∼360–370 cm⁻¹. An even weaker band is resolved at ∼420 cm⁻¹. At first glance the ⁵⁷Fe-PVDOS profile of 1 appears to be a completely different spectrum compared to the profiles of [(n-Bu)₄N]₂[Fe₄S₄(SPh)₄]³⁴ and D14C mutant of ferredoxin from Pyrococcus furiosus (Pf Fd) shown in Fig. 1b.¹⁹ There seems little resemblance to the ⁵⁷Fe-PVDOS of [4Fe–4S] clusters, which have strong in-core Fe–S stretching bands very approximately in the ∼270–290 and ∼350–380 cm⁻¹ regions (Fig. 1b). However, a simple binary harmonic oscillator model for ⁵⁷Fe–¹²⁵Te vs.⁵⁷Fe–³²S vibrations, assuming the same force constant, predicts a frequency downshift factor of 0.72 because of the large ¹²⁵Te mass. This model yields predicted Fe–Te bands in the ∼190–210 cm⁻¹ and ∼250–270 cm⁻¹ regions, in approximate agreement with the observed ⁵⁷Fe spectrum of 1 (Fig. 1a). Thus, the primary cause of the spectral changes can be attributed to the dramatic change in the chalcogen mass.

Fig. 1 (a) ¹²⁵Te- (red) and ⁵⁷Fe- (blue) PVDOS spectra of (Et₄N)₃[⁵⁷Fe₄¹²⁵Te₄(SPh)₄] = 1 in comparison to (b) ⁵⁷Fe-PVDOS profiles of [(n-Bu)₄N]₂[Fe₄S₄(SPh)₄] (black) and the oxidized D14C mutant of Pyrococcus furiosus ferredoxin (green).

The ¹²⁵Te-PVDOS spectrum of 1 reproduces the two main features of the ⁵⁷Fe spectrum at ∼190 cm⁻¹ and ∼240 cm⁻¹, but at significantly lower intensity. For ¹²⁵Te, the spectrum shows a global maximum at ∼100 cm⁻¹ within a broad feature with further local maxima between 20 and 120 cm⁻¹. Qualitatively, this indicates relatively little ¹²⁵Te motion in these modes, which are the strongest ⁵⁷Fe features. Even without sophisticated analysis, one can envision the Fe ions moving in a much less mobile matrix of Te ions. In contrast, below 100 cm⁻¹ there are modes that show primarily Te motion.

DFT calculations

Structure. DFT modelling has been applied to optimize the structure of 1 (=2 × Et₄N⁺ + [⁵⁷Fe₄¹²⁵Te₄(SPh)₄]³⁻ here, as explained below) and subsequently generate its normal vibrational modes for NRVS simulation. Based on the present X-ray determination, the model included 2 × Et₄N⁺ counter ions available from the unit cell as shown in Fig. 2 and S2,† with one of them (Et₄N_B⁺) moderately dislocated in the absence of the crystal framework; as detailed later in the text, either inclusion or complete exclusion of counter ions provides qualitatively similar predicted ¹²⁵Te/⁵⁷Fe-PVDOS spectra. The [Fe₄Te₄(SPh)₄]³⁻ molecular fragment is adequately reproduced by the DFT structure, including the positioning of its phenyl rings. The average Fe–Te internuclear distance, the key structural parameter in the [4Fe–4Te]¹⁺ core, was reproduced by DFT with a deviation of 0.02 Å only, and the average non-bonding Fe⋯Fe/Te⋯Te distances were respectively under-/over-estimated by ∼0.1 Å (Table S2†). Fine structural details of iron-chalcogen cuboids are known to be defined by an interplay of their (i) electronic properties, (ii) ligand identities and their topology, and (iii) effects from the environment, as exemplified by variability in the S = 1/2 and 3/2 [Fe₄X₄(SR)₄]³⁻ counterparts of 1.^32,36–38 Earlier characterization of 1 indicated the elongated D_2d symmetry of its core,³² where the elongation is exhibited as 4 ‘long’ Fe–Te distances (Fig. 3a). Indeed, the present X-ray/DFT structures can be fit to idealized D_2d cuboids with respective RMSD values of 0.06/0.08 Å as shown in Fig. S3a.† However, careful examination of both the X-ray/DFT structures reveals a division of most noticeably the Te⋯Te and the Fe⋯Fe distances into three subsets each (Fig. 3b and c), consistent with lowering of the D_2d symmetry to the D₂ point group, now with even smaller RMSD values of 0.01/0.03 Å as shown in Fig. S3b.† The electronic state of an [4Fe–4X] core is defined by the [4Fe] 3d electronic configuration, where the Fe sites are generally high-spin. Similarly to its iron–sulfur analogue and following the broken-symmetry (BS) formalism,³⁵ the configuration in 1 can be described as internally ferromagnetic mixed-valence [2Fe^2.5+↑] (S = 9/2) and ferrous [2Fe²⁺↓] (S = 4) pairs, coupled in turn antiferromagnetically (↑↓). We find that localization of the ferromagnetic pairs at sites [Fe₁, Fe₄] and [Fe₂, Fe₃] (Fig. 2 and 3b) in a BS-DFT solution having total M_S = 1/2 leads to a description of 1 coherent with the fine distribution of internuclear distances in the X-ray data. Indeed, (i) the spin-dependent delocalization³⁸ favoring ferromagnetic coupling is most feasible in the ‘short’ Fe₁⋯Fe₄ and Fe₂⋯Fe₃ interactions as seen in the X-ray structure. Interestingly, the optimized Fe⋯Fe distances in the [Fe₁, Fe₄] and [Fe₂, Fe₃] pairs in 1 are essentially invariant with respect to the identity of either the mixed-valence (excess-spin, ↑) or ferrous (↓) pair, unlike e.g. the [4Fe–4S]⁺ cluster from the nitrogenase Fe protein developing a ∼0.1 Å shorter contact in its distinct [2Fe^2.5+↑] pair.²⁵ In contrast, (ii) the ‘long’ Fe₁⋯Fe₂ and Fe₃⋯Fe₄ distances are favored by steric repulsion between the in-plane phenyl rings (Fig. 2 and S2†); from DFT, alternative localization of the ferromagnetic Fe sites in the [Fe₁, Fe₂] and [Fe₃, Fe₄] pairs leads to ∼5 kcal mol⁻¹ higher energy and a core structure which does not map well with its X-ray reference. Finally, (iii) the two Et₄N⁺ counter ions play a role in the D_2d-to-D₂ removal of degeneracy by means of their sandwich-like interaction with the [4Fe–4Te] core; a similar effect can possibly be achieved by specific crystal packing.

Fig. 2 Overlay of the X-ray (wire) and DFT-optimized (ball-and-stick) structures of 1, combined from the [Fe₄S₄(SPh)₄]³⁻ and 2 × Et₄N⁺ molecular fragments. 3 × C₂ two-fold axes of the idealized D₂ symmetry of the [4Fe–4Te]⁺ core are shown. Red and green bubbles around the numbered Fe sites correspond to the positive (↑) and negative (↓) 0.01 a.u. point spin density isosurfaces from the representative DFT solution. Hydrogen atoms are omitted for clarity. For extra details, see ESI Fig. S2–S4.†

Fig. 3 Alternative representations of the iron–tellurium cuboidal core of 1 as (a) an [4Fe–4Te] hexahedron showing 12 × Fe–Te bonding interactions, and (b) a compound of smaller [4Fe] and larger [4Te] tetrahedra showing 6 × Fe⋯Fe and 6 × Te⋯Te nonbonding interactions. The internuclear distances are characterized in subsets of ‘long’, ‘intermediate’, and (optionally) ‘short’, with (c) their mean values in the respective polyhedra given as found in the X-ray/DFT-optimized structures. For a detailed account of these distances, see ESI Table S2.† Broken-symmetry (BS) spin alignment (↑ or ↓) of the [4Fe] sites applied in the DFT solution is shown in (b).

In order to compare the [4Fe–4Te]⁺ species to its far more common [4Fe–4S]⁺ counterpart, we likewise analyzed computationally a 1-S species, where the 4 × Te nuclei of 1 were substituted by 4 × S. This analysis included re-optimization of the 1-S model, resulting in a more compressed core structure due to the smaller bridging sulfide dimensions (Fig. S4†). In line with earlier X-ray determinations,³² the mean DFT-optimized S⋯S distance in 1-S is strikingly ∼0.7 Å shorter than the corresponding Te⋯Te distance in 1, and the Fe⋯Fe distances are those least affected (Table S2†). The recognized ‘stereochemical softness’³⁶ of the reduced [4Fe–4X]⁺ core is manifested here as a modified distribution of the Fe–X distances when X is either Te (1) or S (1-S), yet retaining the matching cuboid elongation depicted vertically in Fig. 3a.

Spectra. The DFT-calculated ⁵⁷Fe- and ¹²⁵Te-PVDOS both follow well the observed NRVS spectra of 1, having deviations within 15 cm⁻¹ between the major observed and calculated band positions as displayed in Fig. 4a and b and S5.† Rigorously unavailable from the experiment, a ⁵⁷Fe/¹²⁵Te ≈ 1.1 ratio between the two PVDOS integral intensities is predicted, implying approximately an equal relative distribution of the [4Fe–4Te]¹⁺ vibrational energy between the cationic 4 × Fe and anionic 4 × Te sites. The PVDOS profiles indicate that the vibrational energy of the lighter ⁵⁷Fe nuclei is concentrated mostly within ∼180–260 cm⁻¹ (Fig. 4a), and that of the heavier ¹²⁵Te nuclei mostly within ∼50–150 cm⁻¹ (Fig. 4b). The two regions are non-overlapping; the coupling between the 4 × Fe and 4 × Te motion is therefore relatively weak. As mentioned in the previous section, NRVS-based ⁵⁷Fe/¹²⁵Te-PVDOS nevertheless display complementary bands implying correlated Fe and Te vibrations at, respectively, 191/193 and 241/242 cm⁻¹; the DFT-based ⁵⁷Fe/¹²⁵Te-PVDOS reproduces these features at 186/185 and 255/253 cm⁻¹ (Fig. 4a and b).

Fig. 4 (a) ⁵⁷Fe- and (b) ¹²⁵Te-PVDOS spectra of 1 from NRVS measurements and DFT calculations, followed by (c) DFT-based KED spectra of the relative nuclei motion in bonding 12 × Fe–Te and 4 × Fe–S_Ph, and non-bonding 6 × Fe⋯Fe and 6 × Te⋯Te interactions. Matching NRVS and DFT bands are labelled with their positions. Non-broadened (individual normal mode) DFT PVDOS intensities are additionally provided in a stick-style. For the sticks labelled with dots, corresponding mode animations are available as part of the ESI;† for the four dots circled in black, corresponding modes are depicted in Fig. 5.

A smaller DFT model 1′ = [Fe₄Te₄(SPh)₄]³⁻, which lacks the Et₄N⁺ counter ions, has been considered additionally. Both the more complete 1 and the size-reduced 1′ models produce qualitatively similar PVDOS spectra as shown in Fig. S5.† Inclusion of the counter ions in 1 (vs. exclusion in 1′), however, provides a more realistic distribution of the intensities in the 180–260 cm⁻¹ range of the ⁵⁷Fe-PVDOS, and as well a better population of the low-frequency <100 cm⁻¹ portion of the ¹²⁵Te-PVDOS.

DFT-based kinetic energy distribution (KED) spectra shown in Fig. 4c facilitate analysis of the normal modes in terms of relative displacements in the nuclei pairs. The calculated ⁵⁷Fe-PVDOS spectrum is largely reproduced by the Fe–Te KED profile, implying that the Fe nuclei vibrate in a predominantly static framework of the Te nuclei. The ¹²⁵Te-PVDOS is in contrast similar to the ‘Te-only’ Te⋯Te profile, as if the Te nuclei moved independently. ⁵⁷Fe nuclei vibrations in the stronger Fe–S_Ph (vs. weaker Fe–Te) bonds produce the two NRVS/DFT high-end bands at 371/361 and 424/412 cm⁻¹, respectively (Fig. 4a).

Breakdown of the calculated PVDOS into contributions from individual normal modes, see Fig. 4a and b and 5, provide further characterization of 1. Of remarkable note is a mode calculated at 148 cm⁻¹ and having prominent ¹²⁵Te-PVDOS (and Te⋯Te KED) intensity, yet vanishingly weak in ⁵⁷Fe-PVDOS. Centered in the broad ∼50 cm⁻¹ region devoid of other bands, the 148 cm⁻¹ mode directly corresponds to the experimental ¹²⁵Te band observed at 154 cm⁻¹. DFT reveals this as a pseudo-A₁^Te (totally-symmetric) character ‘breathing’ of the [4Te] tetrahedron depicted in Fig. 5b, a type of vibration well-known from studies on [4Fe–4S] and other cubic clusters.^19,34,39

Fig. 5 Arrow-style representations of selected calculated normal modes in the [4Fe–4Te]⁺ core of 1, having (a) Te⋯Te, (b) ‘breathing’ ∼A₁^Te, (c) Fe–Te, and (d) Fe⋯Fe character. The arrow sizes are proportional to the nuclei displacement amplitudes; actual amplitudes in (b) are respectively ∼0.02/0.005 Å for Te/Fe. Animations for these and several other normal modes are provided as part of the ESI.†

To disentangle the (i) structural/electronic properties and (ii) nuclear masses in the effects determining the PVDOS, we introduce two additional DFT systems 1-³²Te_f and 1-¹²⁵S_f having fictitious isotope nuclei with the ‘native’ Te and S masses interchanged, yet their structures correspondingly equal to 1 = 1-¹²⁵Te and 1-S = 1-³²S. As seen from the PVDOS comparison in Fig. 6, the 1-¹²⁵Te and 1-¹²⁵S_f systems, despite their elementary (and thus chemically) different content and different structures, display very similar pairs of both ⁵⁷Fe- and ¹²⁵Te(¹²⁵S)-PVDOS profiles; a blue-shift of ∼15–20 cm⁻¹ in the 1-¹²⁵S_fvs.1-¹²⁵Te bands is explained by the shorter Fe–S vs. Fe–Te distances (Fig. S4 and Table S2†). In contrast, the variation is more prominent between the 1-³²S and 1-³²Te_f DFT spectra, particularly in the ³²S(³²Te)-PVDOS. The ⁵⁷Fe-PVDOS intensities in the 1-³²S(³²Te_f) systems become more evenly distributed over the entire spectral range vs. sharper bands of 1-¹²⁵Te(¹²⁵S_f) consolidated around 180–280 cm⁻¹, as seen in Fig. 6a. Displaying similarities to the [4Fe–4S] NRVS spectra observed earlier,^19,34 see spectra in Fig. 1b, this behaviour of 1-³²S(³²Te_f) is explained by a stronger vibrational coupling between the motion of Fe and the lighter 32-mass (vs. heavier 125-mass) nuclei in the cubic core. The naturally relevant 1-³²S system shows the strongest coupling, producing ⁵⁷Fe- (Fig. 6a) and ³²S-PVDOS (Fig. 6b) profiles that nearly follow each other.

Fig. 6 DFT-simulated (a) ⁵⁷Fe- and (b) ^125/32Te/S-PVDOS spectra from models 1 (=1-¹²⁵Te) and 1-S (=1-³²S), as well as their fictitious ³²Te- and ¹²⁵S-isotopologues 1-³²Te_f and 1-¹²⁵S_f.

Redox-dependent properties. The synthetic [4Fe–4S] phenylthiolate-coordinated analogue to the [4Fe–4Te] species 1 has been characterized by ⁵⁷Fe-NRVS earlier, as shown in Fig. 1b,³⁴ albeit in the oxidized [4Fe–4S]²⁺ state (in contrast to the above described 1-S model that has the reduced [4Fe–4S]¹⁺ core); adding credibility to the present DFT methodology, the calculated spectrum of the oxidized model [1-S]⁺ compares well to the experimental line shape as shown in Fig. S6a.†

The ⁵⁷Fe-PVDOS comparison drawn above for the reduced 1vs.1-S DFT models (Fig. 6a) applies as well to their 1e⁻ oxidized [1]⁺vs. [1-S]⁺ counterparts (Fig. S6a vs. b†). Both Te/S chalcogenide alternatives display only minor (Te-variant) to moderate (S-variant) redox-dependent alterations in their spectra, where the S-variant behaves similar to the [4Fe–4S]^+/2+ cofactor from ferredoxin characterized by ⁵⁷Fe-NRVS earlier.^19,20 The major bands display blue shifts within 10 cm⁻¹ upon oxidation, explained by +0.8/+1.5% larger volume of respectively the [4Fe–4Te/S]¹⁺vs. [4Fe–4Te/S]²⁺ cores (Table S3†). While these redox-dependent changes can be considered as insignificant, relative trends in the ⁵⁷Fe-PVDOS and core volume adjustments indicate that the S-to-Te substitution may present a cubic cluster having enhanced proficiency in electron transfer. This expectation is now supported by a significant ∼1.5 ratio (S/Te-variant) in the calculated reorganization energies of structural relaxation following electron transfer for the 1-S/[1-S]⁺ (17.4 kJ mol⁻¹) vs.1/[1]⁺ (11.6 kJ mol⁻¹) redox couples (Table S4†). The [4Fe–4Te] core with its volume ∼1.4 times larger than that of [4Fe–4S] (Table S3†) is therefore expected to be more efficient in charge delocalization.

Conclusions

Iron–sulfur clusters play many important roles in biological systems, ranging from purely structural functions to electron transfer, small molecule sensing, and as components of active sites. However, the large amount of sulfur in most biological samples poses a challenge for spectroscopists who want to investigate these clusters from the sulfur point of view.

We have shown here that sulfur to tellurium exchange, combined with NRVS measurements from both ⁵⁷Fe and ¹²⁵Te points of view, creates a powerful tool for chalcogenide-specific vibrational spectroscopy. Additionally, DFT calculations reproduced and rationalized the obtained structure and spectra. Our results clearly demonstrate corresponding bands in ⁵⁷Fe- and ¹²⁵Te-NRVS spectra that could successfully be assigned to the same normal modes. Unlike [4Fe–4S] clusters where the vibrational coupling of the Fe and S motion is tight, a significant difference in the Fe vs. Te nuclei masses and slightly weaker Fe–Te force constants in the [4Fe–4Te] cuboid both lead to divergence in the vibrational signatures of the [4Fe] and [4Te] interposed tetrahedra. Herein, this exchange leads to clearly isolated Fe-only motions displaying sharper bands within the ⁵⁷Fe spectrum by almost cancelling out the Fe–Te vibrational coupling. Thus, (selective) Te incorporation can be potentially used to simplify complex ⁵⁷Fe-NRVS spectra of proteins containing multiple Fe–S clusters, and to allow for a more discrete discussion on the properties of such active centers.

This work is the first ¹²⁵Te-NRVS investigation of an isolated [4Fe–4Te] cluster, and it lays the basis for bioinorganic spectroscopic investigation of Te substitution. Our disclosure of the [4Te]-‘breathing’ mode in 1 is an example of finding by ¹²⁵Te-NRVS which would evade determination by ⁵⁷Fe-NRVS which is commonly used nowadays. Se substitution for S in biological [2Fe–2S]⁴⁰ and [4Fe–4S]⁴¹ clusters have been known for decades,⁴² and Te substitution in various electron-transport proteins as well as enzyme active sites should be feasible. For example, a [4Fe–4Se] cluster was shown to replace the [4Fe–4S]_H sub-cluster in [FeFe]-hydrogenase with full retention of activity.⁴³ [4Fe–4Te] clusters could presumably be incorporated into hydrogenase in the same manner and studied by ¹²⁵Te-NRVS. Since Se can also be incorporated into thiolate positions of the H-cluster azadithiolate ligand,⁴⁴ comparable experiments using Te and ¹²⁵Te-NRVS can also be envisaged. In small molecules, Te can be a replacement for terminal thiolate ligands or thioethers, although it remains to be seen whether tellurocysteine or telluromethionine can be inserted into proteins in the manner already established for selenocysteine. Finally, Se has been shown to incorporate into the nitrogenase MoFe cofactor,⁴⁵ and the analogous experiments with Te might be interesting. Although it has a larger ionic radius and lower electronegativity, tellurium often exhibits chemistry similar to sulfur, and it can replace sulfur in many structures. Our calculations for the [4Fe–4Te] vs. [4Fe–4S] clusters indicate that, while iron–tellurium cofactors are bigger in their size, they have lower reorganization energies favorable during events of electron transfer. These results clearly showcase the potential of ¹²⁵Te-NRVS in such modified systems.

Conflicts of interest

There are no conflicts of interest to declare.

Acknowledgements

U.-P. A. acknowledges funding by the Fonds der Chemischen Industrie (Liebig Grant), the Deutsche Forschungsgemeinschaft (Emmy Noether Grant AP242/2-1), and the Fraunhofer Internal Programs (Grant ATTRACT 097-602175). F. W. thanks the Studienstiftung des Deutschen Volkes for a PhD fellowship. N. M., H. W., and S. P. C. acknowledge support by the National Institutes of Health (grant number GM-65440) and the Einstein Foundation Berlin (grant number EVF-2016-277). V. P. and M. K. acknowledge funding by the Deutsche Forschungsgemeinschaft under Germany's Excellence Strategy – EXC 2008/1 (UniSysCat) – 390540038. We acknowledge DESY (Hamburg, Germany), a member of the Helmholtz Association HGF, for the provision of experimental facilities. The ¹²⁵Te-NRVS work was carried out at PETRA III and we would like to thank I. Sergeev and O. Leupold for assistance in using the P01 beamline. The ⁵⁷Fe-NRVS work was performed at SPring-8 under proposals 2018B0141 and 2018A1409.

Notes and references

1. K. F. Hsu, S. Loo, F. Guo, W. Chen, J. S. Dyck, C. Uher, T. Hogan, E. K. Polychroniadis and M. G. Kanatzidis, Science, 2004, 303, 818–821.
2. Y. Xiao and L.-D. Zhao, npj Quantum Mater., 2018, 3, 55.
3. N. Harmgarth, F. Zörner, P. Liebing, E. P. Burte, M. Silinskas, F. Engelhardt, F. T. Edelmann and Z. Für, Z. Anorg. Allg. Chem., 2017, 643, 1150–1166.
4. J. Wen, G. Xu, G. Gu, J. M. Tranquada and R. J. Birgeneau, Rep. Prog. Phys., 2011, 74, 124503.
5. Z. Xu, J. A. Schneeloch, J. Wen, B. L. Winn, G. E. Granroth, Y. Zhao, G. Gu, I. Zaliznyak, J. M. Tranquada, R. J. Birgeneau and G. Xu, Phys. Rev. B: Condens. Matter Mater. Phys., 2017, 96, 134505.
6. L. A. Ba, M. Döring, V. Jamier and C. Jacob, Org. Biomol. Chem., 2010, 8, 4203–4216.
7. T. C. Stadtman, Annu. Rev. Biochem., 1996, 65, 83–100.
8. A. Böck, K. Forchhammer, J. Heider and C. Baron, Trends Biochem. Sci., 1991, 16, 463–467.
9. B. L. Cohen, Geochim. Cosmochim. Acta, 1984, 48, 203–205.
10. S. E. Ramadan, A. A. Razak, A. M. Ragab and M. El-Meleigy, Biol. Trace Elem. Res., 1989, 20, 225–232.
11. B. B. Buchanan and D. I. Arnon, Adv. Enzymol., 1970, 33, 119–176.
12. W. Lubitz, H. Ogata, O. Rüdiger and E. Reijerse, Chem. Rev., 2014, 114, 4081–4148.
13. J. C. Crack, J. Munnoch, E. L. Dodd, F. Knowles, M. M. Al Bassam, S. Kamali, A. A. Holland, S. P. Cramer, C. J. Hamilton, M. K. Johnson, A. J. Thomson, M. I. Hutchings and N. E. Le Brun, J. Biol. Chem., 2015, 290, 12689–12704.
14. H. Beinert, JBIC, J. Biol. Inorg. Chem., 2000, 5, 2–15.
15. J. B. Broderick, B. R. Duffus, K. S. Duschene and E. M. Shepard, Chem. Rev., 2014, 114, 4229–4317.
16. N. D. Lanz and S. J. Booker, Biochim. Biophys. Acta, Mol. Cell Res., 2015, 1853, 1316–1334.
17. K. Yokoyama and E. A. Lilla, Nat. Prod. Rep., 2018, 35, 660–694.
18. W. Zeng, N. J. Silvernail, W. R. Scheidt and J. T. Sage, in Encyclopedia of Inorganic and Bioinorganic Chemistry, ed. R. A. Scott, John Wiley & Sons, Ltd, Chichester, UK, 2011.
19. D. Mitra, V. Pelmenschikov, Y. Guo, D. A. Case, H. Wang, W. Dong, M.-L. Tan, T. Ichiye, F. E. Jenney, M. W. W. Adams, Y. Yoda, J. Zhao and S. P. Cramer, Biochemistry, 2011, 50, 5220–5235.
20. L. Lauterbach, L. B. Gee, V. Pelmenschikov, F. E. Jenney, S. Kamali, Y. Yoda, M. W. W. Adams and S. P. Cramer, Dalton Trans., 2016, 45, 7215–7219.
21. W. R. Scheidt, J. Li and J. T. Sage, Chem. Rev., 2017, 117, 12532–12563.
22. E. J. Reijerse, C. C. Pham, V. Pelmenschikov, R. Gilbert-Wilson, A. Adamska-Venkatesh, J. F. Siebel, L. B. Gee, Y. Yoda, K. Tamasaku, W. Lubitz, T. B. Rauchfuss and S. P. Cramer, J. Am. Chem. Soc., 2017, 139, 4306–4309.
23. V. Pelmenschikov, J. A. Birrell, C. C. Pham, N. Mishra, H. Wang, C. Sommer, E. Reijerse, C. P. Richers, K. Tamasaku, Y. Yoda, T. B. Rauchfuss, W. Lubitz and S. P. Cramer, J. Am. Chem. Soc., 2017, 139, 16894–16902.
24. K. D. Sutherlin, B. S. Rivard, L. H. Böttger, L. V. Liu, M. S. Rogers, M. Srnec, K. Park, Y. Yoda, S. Kitao, Y. Kobayashi, M. Saito, M. Seto, M. Hu, J. Zhao, J. D. Lipscomb and E. I. Solomon, J. Am. Chem. Soc., 2018, 140, 5544–5559.
25. D. Mitra, S. J. George, Y. Guo, S. Kamali, S. Keable, J. W. Peters, V. Pelmenschikov, D. A. Case and S. P. Cramer, J. Am. Chem. Soc., 2013, 135, 2530–2543.
26. H.-C. Wille, R. P. Hermann, I. Sergueev, U. Pelzer, A. Möchel, T. Claudio, J. Perßon, R. Rüffer, A. Said and Yu. V. Shvyd’ko, EPL, 2010, 91, 62001.
27. I. Sergueev, H.-C. Wille, R. P. Hermann, D. Bessas, Yu. V. Shvyd’ko, M. Zając and R. Rüffer, J. Synchrotron Radiat., 2011, 18, 802–810.
28. D. Bessas, I. Sergueev, H.-C. Wille, J. Perßon, D. Ebling and R. P. Hermann, Phys. Rev. B: Condens. Matter Mater. Phys., 2012, 86, 224301.
29. D. Bessas, W. Töllner, Z. Aabdin, N. Peranio, I. Sergueev, H.-C. Wille, O. Eibl, K. Nielsch and R. P. Hermann, Nanoscale, 2013, 5, 10629–10635.
30. R. E. Simon, I. Sergueev, I. Kantor, A. Kantor, J. Perßon and R. P. Hermann, Semicond. Sci. Technol., 2014, 29, 124001.
31. P. Bauer Pereira, I. Sergueev, S. Gorsse, J. Dadda, E. Müller and R. P. Hermann, Phys. Status Solidi B, 2013, 250, 1300–1307.
32. P. Barbaro, A. Bencini, I. Bertini, F. Briganti and S. Midollini, J. Am. Chem. Soc., 1990, 112, 7238–7246.
33. H.-O. Stephen, C. Chen, G. Henkel, K. Griesar and W. Haase, J. Chem. Soc., Chem. Commun., 1993, 886–888.
34. Y. Xiao, M. Koutmos, D. A. Case, D. Coucouvanis, H. Wang and S. P. Cramer, Dalton Trans., 2006, 2192–2201.
35. L. Noodleman and D. A. Case, in Advances in Inorganic Chemistry, Elsevier, 1992, vol. 38, pp. 423–470.
36. M. J. Carney, G. C. Papaefthymiou, K. Spartalian, R. B. Frankel and R. H. Holm, J. Am. Chem. Soc., 1988, 110, 6084–6095.
37. P. Venkateswara Rao and R. H. Holm, Chem. Rev., 2004, 104, 527–560.
38. P. W. Anderson and H. Hasegawa, Phys. Rev., 1955, 100, 675–681.
39. R. S. Czernuszewicz, K. A. Macor, M. K. Johnson, A. Gewirth and T. G. Spiro, J. Am. Chem. Soc., 1987, 109, 7178–7187.
40. J. C. M. Tsibris, M. J. Namtvedt and I. C. Gunsalus, Biochem. Biophys. Res. Commun., 1968, 30, 323–327.
41. J. Meyer and J.-M. Moulis, Biochem. Biophys. Res. Commun., 1981, 103, 667–673.
42. J. Meyer, J.-M. Moulis, J. Gaillard and M. Lutz, in Advances in Inorganic Chemistry, Elsevier, 1992, vol. 38, pp. 73–115.
43. J. Noth, J. Esselborn, J. Güldenhaupt, A. Brünje, A. Sawyer, U.-P. Apfel, K. Gerwert, E. Hofmann, M. Winkler and T. Happe, Angew. Chem., Int. Ed., 2016, 55, 8396–8400.
44. L. Kertess, F. Wittkamp, C. Sommer, J. Esselborn, O. Rüdiger, E. J. Reijerse, E. Hofmann, W. Lubitz, M. Winkler, T. Happe and U.-P. Apfel, Dalton Trans., 2017, 46, 16947–16958.
45. T. Spatzal, K. A. Perez, J. B. Howard and D. C. Rees, eLife, 2015, 4, e11620.

---

Footnotes

† Electronic supplementary information (ESI) available: Supplementary Tables S1–S4, Fig. S1–S6, Materials and methods, atomic Cartesian coordinates, and animated vibrational modes. CCDC 1908709. For ESI and crystallographic data in CIF or other electronic format see DOI: 10.1039/c9sc02025j

‡ These authors contributed equally.

---