Simultaneous Speciation, Structure, and Equilibrium Constant

Nov 8, 2017 - ... both the structural variations and stability constants of the forming complexes were determined from the same measurement series, pr...
3 downloads 12 Views 1MB Size
Download PDF
Article pubs.acs.org/IC

Cite This: Inorg. Chem. XXXX, XXX, XXX-XXX

Simultaneous Speciation, Structure, and Equilibrium Constant Determination in the Ni2+−EDTA−CN− Ternary System via HighResolution Laboratory X‑ray Absorption Fine Structure Spectroscopy and Theoretical Calculations Éva G. Bajnóczi,* Zoltán Németh, and György Vankó Wigner Research Centre for Physics, Hungarian Academy of Sciences, Budapest, Hungary S Supporting Information *

ABSTRACT: Even quite simple chemical systems can involve many components and chemical states, and sometimes it can be very difficult to differentiate them by their hardly separable physical−chemical properties. The NiII−EDTA−CN− (EDTA = ethylenediaminetetraacetic acid) ternary system is a good example for this problem where, in spite of its fairly simple components and numerous investigations, several molecular combinations can exist, all of them not having been identified unambiguously yet. In order to achieve a detailed understanding of the reaction steps and chemical equilibria, methods are required in which the structural transitions in the different reaction steps can be followed via element-selective complex spectral feature sets. With the help of our recently developed von Hámos type high-resolution laboratory X-ray absorption spectrometer, both the structural variations and stability constants of the forming complexes were determined from the same measurement series, proving that X-ray absorption spectroscopy can be considered as a multifaced, table-top tool in coordination chemistry. Furthermore, with the help of theoretical calculations, independent structural evidence was also given for the formation of the [NiEDTA(CN)]3− mixed complex.

■

K1

INTRODUCTION Coordination chemists are commonly interested not only in the stability but also in the structure of the forming complex(es) in a solution phase. This structure determination can often be challenging because most conventional coordination chemistry tools, such as potentiometry, conductometry, UV−vis and fluorescence spectroscopy, give no or limited information about the structure of the complexes; however, the stability constants can be derived from these measurements with high accuracy. The different structure determination methods usually require either crystalline or at least solid material (e.g., X-ray diffractometry) or concentrations and/or temperatures (e.g., large-angle X-ray scattering and electron spin resonance spectroscopy) different from the ones used for the stability measurements; therefore, the obtained structures are not necessarily the same among the different conditions. In addition, the physicochemical properties of the consecutively forming complexes can vary within a few orders of magnitude (molar absorbance, conductivity, etc.), or for optical spectroscopies, the overlapping absorption bands cannot be distinguished properly, which complicates even the speciation of these solutions. An excellent example for this is the NiII−EDTA−CN− (EDTA = ethylenediaminetetraacetic acid) ternary system, where the following equilibrium reactions should be considered: © XXXX American Chemical Society

Ni 2 +(aq) + EDTA4 − ⇌ [NiEDTA]2 −

(1)

K2

[NiEDTA]2 − + CN− ⇌ [NiEDTA(CN)]3 −

(2)

β

[NiEDTA]2 − + 4CN− ⇌ [Ni(CN)4 ]2 − + EDTA4 −

(3)

K4

[Ni(CN)4 ]2 − + CN− ⇌ [Ni(CN)5 ]3 −

(4) 2−

The absorbance spectra of the [NiEDTA] and [NiEDTA(CN)]3− complexes are quite similar,1 the molar absorbance of [Ni(CN)4]2− is an order of magnitude higher, and the absorption band around 450 nm, which belongs to the [Ni(CN)5]3− complex uniquely, is again partly masked by the overlap with the tails of intense UV bands at higher concentrations. 2 Although the stability constants were determined previously for all of the above-mentioned complexes, these values are somewhat ambiguous,1−5 and some of them are indirect, especially the one for the [NiEDTA(CN)]3− mixed complex, because its existence was only proven by kinetic measurements.1,3−5 The existence of further intermediate mixed complexes, like [NiEDTA(CN)2]4− and [NiEDTA(CN)3]5−, is also proposed based on these Received: September 13, 2017

A

DOI: 10.1021/acs.inorgchem.7b02311 Inorg. Chem. XXXX, XXX, XXX−XXX

Article

Inorganic Chemistry

Although, upon the addition of KCN, the pH of the solution increased, the concentration distribution in the chemical equilibrium given in eqs 1−4 is not affected by the hydroxide concentration in this range. A concentrated (∼20 M) NaOH solution was prepared to remove the carbonate content of the solution, and the exact concentration was determined by measuring its density at 25 °C.17,18 First, NiCl2·2H2O was dissolved in a small amount of water, and then the EDTA solution (the required mass of Na2EDTA·2H2O, dissolved in a stoichiometric volume of 1 M NaOH to reach full deprotonation of EDTA) was added; finally, the pH was set, and the solution was added to a volumetric flask. EDTA was used in excess to avoid the formation of Ni(OH)2 during the preparation. The samples for the X-ray measurements were prepared by adding the required amount of KCN in solid form to 5 cm3 of the stock solution, and they were prepared several hours before measurement to make sure that the equilibrium was reached. (In fact, the time scales of the processes were verified in an 11-day-long experimental run, where the spectral variations were recorded at different [NiEDTA] 2− and CN − concentrations.) For all samples used in the XANES measurements, the solutions contained c(NiCl2) = 0.25 M and c(EDTA) = 0.30 M and the pH was set to 11. The concentrations of KCN in the different solutions were 0, 0.25, 0.50, 0.75, 1.00, 2.00, and 3.00 M. Because NiII was added as NiCl2, formation of the [NiCl]+ complex might also play a role, but because of its small stability constant compared to the other coexisting complexes, it can be neglected, as model calculations have proven. The role of nickel(II) hydroxido complexes and/or precipitation does not need to be taken into account either because all experiments were done at an excess of EDTA. Important Safety Note. As is well-known, CN− ions form highly volatile and toxic HCN in acidic media; therefore, the pH of the stock solution should be kept at least around 11 to avoid the formation of any toxic gases. For the same reason, all of the preparations have to be performed under a properly functioning fume hood using appropriate safety gloves. The sample holder used for the laboratory XANES measurements was airtight; thus, there was no risk for any poisoning during spectral collection. UV−Vis Spectra. The UV−vis spectra were collected with the following setup in the transmittance geometry in a 1-cm-path-length quartz cuvette: AvaLight-DHc light source, AvaSpec-ULS3648 spectrophotometer, and FC-UV200-2 fiber optics. The concentration was set to not exceed the maximum absorbance value of 1. Experimental XANES. The laboratory XANES measurements were performed with a newly developed table-top spectrometer.14 The samples were in a custom-made liquid cell consisting of two stainless steel pieces, which could be bolted via a 1 mm steepness screw, so that the sample thickness could be varied between 0 and 1 mm. The 20mm-diameter windows were covered by 25 μm Kapton foil, transmitting more than 96% of the radiation in the investigated energy range. The two holes on the top and bottom of the sample holder allowed us to fill the cell without any bubbles. During the measurements, the fluid cell was placed right in front of the X-ray source in a fixed position with a constant 0.5 mm sample thickness. The voltage and current of the X-ray tube were set to 9 kV and 40 mA, respectively. The n = 4 reflection of a Si(111) analyzer crystal was used to get the energy-resolved XANES spectrum on the position-sensitive Dectris Mythen strip detector with a threshold of 8 keV. The data were accumulated for 1 h per sample; thus, the total acquisition time of the experiment was within one working day. The raw spectrum was constructed as a common X-ray absorbance spectrum: A = log(I0/I), where I is the spectrum measured with the sample in the beam path and I0 is the spectrum measured with distilled water (with the same thickness as the sample) in the beam path. The raw data were normalized with the help of the Athena program (part of the Demeter package).19 The normalized spectra were binned by five points and smoothed with Savitzky−Golay filtering.20 The energy calibration of the instrument was performed with a metallic nickel foil, specially designed for the reference material for Xray absorbance measurements (purchased from EXAFS Materials).

experiments; however, their formation constant is undetermined, and no clear experimental evidence is given for them.1 All of these problems could be overcome with an elementselective tool, like X-ray absorption spectroscopy (XAS), which provides comprehensive information about the structure of the complex(es). The X-ray absorption near-edge structure (XANES; 50−100 eV around the absorption edge) reveals the oxidation and spin states of the metal center as well as the local coordination geometry around it, while the extended Xray absorption fine structure (EXAFS; up to ∼1000 eV above the absorption edge) gives information about the distance of the neighboring atoms and their types.6,7 Because of the great advantage of the method that no crystalline material is required, the structure of the complex can be investigated directly in a solution phase in a wide concentration range because of the different detection techniques. Although XAS is usually considered to be a structure determination method, it can also be used to determine the formation and stability constants when a specific spectral feature, related to a major geometry change, can be assigned to the forming complex(es).8,9 Until now, these experiments required synchrotrons as X-ray sources because only these facilities have a narrow-energy bandwidth, monochromatic radiation, and sufficient brilliance to study the fine variations of the absorption coefficient. Because of the substantial interest in this robust method, the beamlines available for XAS experiments are heavily overbooked (even a successful proposal with a hot topic theme can wait a year to get only a few days of beamtime); this time scale is unacceptably long in contemporary competitive research. As a result of the recent developments in the field of laboratory Xray sources, detectors, and monochromators, however, a few laboratory XAS setups have been developed based on different working principles.10−14 Our newly developed table-top von Hámos X-ray absorption spectrometer is optimized for routine characterization of both solid and solution samples. The spectrometer consists of a conventional X-ray tube with a copper anode, a positionsensitive 1D strip detector, and a segmented, cylindrical silicon single-crystal analyzer,15 arranged in the von Hámos geometry. The spectrometer is quite flexible and quickly adjustable, can be used for most of the transition metals (K- and/or L-edges), contains no scanning components, and is operated in air; thus, the whole setup is simple and easy to run. The detailed description of the spectrometer and some examples on solid samples can be found in ref 14. Here we demonstrate that this spectrometer can be used as a powerful tool in coordination chemistry, even on a daily basis, through the detailed laboratory XAS investigation of the NiII− EDTA−CN− ternary system, where well-defined structural changes are expected (see Figure 1 for the calculated and optimized structures of the possible complexes) and show our efforts in clarifying its speciation via identifying the forming complexes with the help of complementary quantum-chemical calculations.

■

EXPERIMENTAL SECTION

Sample Preparation. Most of the chemicals, NaOH, Na2EDTA· 2H2O, and KCN, were used as received (Reanal, puriss.). NiCl2·6H2O (Reanal, puriss.) was dried in a drying oven at 80 °C for 17 h to get a yellowish, highly hygroscopic NiCl2·2H2O, which was used to prepare the stock solution.16 The stock solution contained 0.25 M NiCl2 and 0.30 M ethylenediaminetetraacetic acid (EDTA), and the pH was set to 11 to avoid the formation of HCN later upon the addition of KCN. B

DOI: 10.1021/acs.inorgchem.7b02311 Inorg. Chem. XXXX, XXX, XXX−XXX

Article

Inorganic Chemistry The pixel number of the first maximum value of the first derivative of the metallic Ni K-edge X-ray absorption spectrum was set to 8333.0 eV, the tabulated value of the Ni K-edge. The energy value of all of the other pixels was calculated according to the Bragg law, taking into account the type of used analyzer crystal [Si(111)], its bending radius (25 cm), the reflection order used (n = 4), and the pixel width of the detector (50 μm). Analysis of Chemical Equilibria. To determine the species participating in the equilibrium process at different concentration levels as well as the corresponding formation constants, the preedge region was used. The intensity of the preedge peak was integrated using splines between 8332 and 8340 eV. The integrated preedge peak intensity value belonging to the solution with a 1:4 NiII/CN− ratio was chosen to be 1, and all of the others were normalized to it. The normalized XANES preedge intensity values were treated as the absorbance analogue quantity in the PSEQUAD program.21 This software is able to carry out any kind of equilibrium computation in aqueous media, so it is able to fit the experimental data to calculate the stability constants of the mixed and pentacyano complexes. The description of the basic theory behind the program given in the next paragraph is based on the reference manual, where further details can also be found.21 In general, an equilibrium system can be described via the massbalance equation: n

Ci =

n

k

∑ αji[Sj] = ∑ αjiβj ∏ [ct ]αjt j=1

complexes with density functional theory (DFT) using the ORCA 3.0 program package.22 In all cases, the TZVP basis set23,24 and conductor-like screening model (COSMO)25 were employed; the latter approximates the solvation effects by embedding the molecule in a polarizable medium with a dielectric constant of water. Several functionals have been tested, including BP86,26,27 B3LYP,28−30 B3LYP*,31 and TPSSH;32,33 a dense integration grid (Grid4) was employed, with a higher integration accuracy at nickel (Grid7). The BP86-optimized structures are shown in Figure 1, and the atomic

(i = 1, ..., k)

t=1

j=1

Figure 1. Calculated and optimized structures for all of the corresponding complexes: (A) [NiEDTA]2−; (B) [NiEDTA(CN)]3−; (C) [Ni(CN)4]2−; (D) [Ni(CN)5]3−.

(5)

where Ci is the total concentration of the ith component, n is the number of species in the system, including the components, Sj is the jth species present in the system, k is the number of components in the system, [ci] is the equilibrium (free) concentration of the ith component, βj is the formation constant of the jth species, βj = [Sj]/∏ki=1[ci]αji, and αji is the stoichiometric number, giving the number of ith components in the jth species. For absorbance-analogue quantities, the measured experimental data (Xl) can be expressed by the free concentration of one or more of the components: n

Xl =

coordinates are given in the Supporting Information; these were used to calculate the XANES spectra with the FEFF9 package.34 The theoretical UV−vis spectra were calculated for all corresponding cases with time-dependent DFT (TD-DFT) at the Tamm−Dancroft approximation (TDA),35,36 and the B3LYP functional was found to provide the best match for the experiments. The calculated UV−vis spectra were broadened by a pseudo-Voigt function [with Lorentz and Gauss full widths at half-maximum (fwhm) of 900 and 1000 cm−1, respectively]. The preedges were also computed within a TD-DFT/ TDA framework, as implemented in ORCA,37,38 using the B3LYP functional. A constant energy shift of +249 eV and a broadening by a pseudo-Voigt function (with Lorentz and Gauss fwhm values of 1.8 and 2 eV, respectively) were employed for all preedge spectra to match the experiment.

k

∑ ϵjlβj ∏ [ci]αji i=1

j=1

(l = 1, 2, ..., p) (6)

where p is the number of wavelengths studied, Xl is the measured absorbance at the lth wavelength, and ϵjl is the molar absorptivity of the jth species at the lth wavelength or energy. PSEQUAD solves eq 5 for the unknown free concentrations and obtains the unknown formation constants and/or molar absorbancies by minimizing the F function: nd

F=

rq

■

RESULTS AND DISCUSSION Spectral Changes. The UV−vis spectra of the [NiEDTA]2− and [Ni(CN)4]2− complexes are displayed in the first and third rows of Figure 2, demonstrating the good agreement between the calculated and experimental spectra, although the theoretical ones are somewhat shifted in wavelength. When the EDTA−CN− mixed and pentacyano complexes cannot be produced at 100%, their molar absorbance spectra cannot be collected. However, their UV−vis spectra can be easily calculated, as shown in the right panels of Figure 2, second and fourth rows. However, the spectra of the [NiEDTA(CN)]3− and [Ni(CN)5]3− complexes cannot be seen directly in the measured spectra (Figure 2, second and fourth rows) because of the masking effects of the spectral features of [NiEDTA]2− or [Ni(CN)4]2−, the other relevant species present in the solution. The pure theoretical spectra for these two cases are therefore only shown in the inset, and in the main panel, this spectrum and the one for the other dominant component are shown, both with increased broadening and a limited wavelength range to compare better with the experiment.

p 2

∑ ∑ ∑ (ΔXl) q=1 i=1

l

(7)

where ΔXl is the difference between the measured and calculated (based on the current model) data, nd is the number of sets of measurements derived from different types of primary experimental data, and rq is the number of experimental points in the qth set of measurements. The stability constants of the [NiEDTA]2− and Ni(CN)42− complexes were held constant during the computations at the following widely accepted values: log K1 = 18.50 for the [NiEDTA]2− and log β = 30.50 for the [Ni(CN)4]2−.3,5 The literature-based distribution was also calculated with these values, together with the average of the values found in the literature of log K2 = 3.6 and log K4 = −0.7 for the mixed and pentacyano complexes.2,3 All of these values resulted from the spectrophotometric determination performed in different concentration ranges for each complex. The simulation of the equilibrium concentrations for optimization of both the experimental total concentrations and the distribution diagram was also performed with the PSEQUAD program.21 Quantum-Chemical Calculations. Molecular geometries were optimized by employing a tight optimization criterion for all C

DOI: 10.1021/acs.inorgchem.7b02311 Inorg. Chem. XXXX, XXX, XXX−XXX

Article

Inorganic Chemistry

The octahedral [NiEDTA]2− and [NiEDTA(CN)]3− complexes have only a weak preedge peak around 8339 eV because the 1s → 3d electronic transitions are dipole-forbidden in this geometry, resulting in a very low cross section (Figure 3A). In contrast, this transition is allowed in the four-coordinated, square-planar [Ni(CN)4]2− and appears as a strong preedge signal at about 8336 eV.39 This dipole transition is allowed in the [Ni(CN)5]3− complex with five-coordinated, squarepyramidal arrangement as well but with reasonably smaller probability and therefore with lower intensity.40 Not only in the preedge but also in the whole XANES region are systematic spectral changes expected, like the slight edge-energy shift as well as the change of the relative intensities and energy positions of the features above the edge (see, e.g., the peaks around 8370 or 8400 eV). Therefore, the XANES spectra can be used as a fingerprint to identify the forming complexes. The experimental Ni K-edge XANES spectra are shown in Figure 3B as a function of the CN− concentration. The XANES spectrum of the CN−-free solution (dark-blue line in Figure 3B) matches well with the calculations corresponding to [NiEDTA]2− with octahedral coordination (dark blue in Figure 3A). When CN− is added to this solution, a new preedge peak appears and grows around 8336 eV, indicating formation of the tetracyano complex. Besides, the K-edge position (determined by the inflection point of the rising edge) shifts by about 3 eV in 0 M ≤ c(CN) ≤ 1 M, and a continuous transition from the octahedral [NiEDTA]2− spectrum to the tetragonal [Ni(CN)4]2− spectrum can be seen in the postedge region (see the Supporting Information), in agreement with the FEFF9 calculations. Because the concentration of CN− exceeds the stoichiometric ratio in Ni(CN)42−, the intensity of the preedge peak starts to decrease, indicating formation of the pentacyano complex, of which the preedge intensity in this region is expected to be significantly lower. Again in agreement with the calculations, above this CN− concentration, the edge position turns back to smaller energies, although the small fraction of Ni(CN)53− and the subtle changes in the postedge spectrum between the tetracyano and pentacyano complexes do not allow further analysis of the postedge region. Identification of the Complexes Formed. The integrated preedge peak area value belonging to the solution with a 1:4 NiII/CN− ratio was chosen to be 1, and all others were normalized to it. These integrated and normalized preedge area values were used as conventional absorbance analogue experimental data in the PSEQUAD fitting process. Because at a 1:4 CN− ratio all of the NiII ions are present in the form of [Ni(CN)4]2−, the mole fraction of NiII in this form is equal to 1; thus, the normalized and integrated preedge area values also can be regarded as the mole fraction of NiII ions in the tetracyano complex form, which is represented by purple dots in Figure 3C. Initially (without any CN−), the entire amount of NiII ions are present as [NiEDTA]2−. With increasing CN− concentration, the ratio of [NiEDTA]2− decreases, while formation of the [Ni(CN)4]2− complex continues until the solution reaches a stoichiometric 1:4 NiII/CN− ratio. However, this increase is not linear, which indicates the presence of a ternary [NiEDTA(CN)]3− complex with octahedral ligand symmetry. The same trend can be seen in the edge shift of the K-edge XANES spectra, while the edge moves toward higher energies monotonously with a cyanide concentration below c(CN−) = 1 M; as expected with the formation of Ni(CN)42−, this shift also lags behind the expected linear change (see Figure S1). This indicates again an intermediate product, for

Figure 2. Experimental and calculated UV−vis spectra of the different NiII−EDTA−CN− complexes. Top row: Spectra of [NiEDTA]2− (blue curves). Second row, left side: Spectrum of a 1:1 [NiEDTA]2−/KCN mixture after 11 h of reaction time (green) compared to the pure NiEDTA spectrum (blue). Second row, right side: Calculated [NiEDTA(CN)]3− complex (green) versus [NiEDTA]2− (blue), both broadened with a Gaussian function to match the experimental data. The inset shows the original calculated spectrum of [NiEDTA(CN)]3−. Third row: Spectra of [Ni(CN)4]2− (orange, label A), where the inset of the calculated spectrum shows the extra peak resulting from the out-of-plane displacement of the Ni2+ ion from the C−N plane (dotted orange, label B). Fourth row, left side: Spectrum of a 1:40 [Ni(CN)4]2−/KCN mixture marking the appearance of the [Ni(CN)5]3− complex (red) compared to the [Ni(CN)4]2− spectrum. Fourth row, right side: Calculated [Ni(CN)4]2− complex (orange) versus [Ni(CN)5]3− (red), both broadened with a Gaussian function to match the experimental data. The inset shows the original calculated spectrum of [Ni(CN)5]3−.

It is also seen from the theoretical spectra that the absorption bands of the four species greatly overlap, especially in the UV region; therefore, their separation based on UV−vis spectroscopy is quite challenging. On the other hand, the complex formation in this ternary system could be easily monitored by XANES spectroscopy via the two main structural changes: while in the [NiEDTA]2− complex, the NiII ion is surrounded by two N and four O atoms in the octahedral geometry, the tetracyano complex is fourcoordinated in a square-planar arrangement, which turns into to a five-coordinated square pyramid upon formation of the pentacyano complex. The calculated XANES and preedge spectra are shown for each species in Figure 3A. D

DOI: 10.1021/acs.inorgchem.7b02311 Inorg. Chem. XXXX, XXX, XXX−XXX

Article

Inorganic Chemistry

Figure 3. (A) XANES spectra calculated by FEFF9 for the [NiEDTA]2−, [NiEDTA(CN)]3−, [Ni(CN)4]2−, and [Ni(CN)5]3− complexes. The inset shows the preedge region for each complex calculated via TD-DFT/TDA. (B) Laboratory XANES spectra of solutions containing constant 0.25 M NiCl2 and 0.30 M EDTA, while the concentration of the added KCN was varied between 0 and 3 M. The inset shows the preedge region with a peak around 8336 eV, representing the formation of four- and/or five-coordinated nickel(II) complexes. (C) Distribution diagram of the following species: [NiEDTA]2− (blue), [NiEDTA(CN)]3− (green), [Ni(CN)4]2− (yellow), and [Ni(CN)5]3− (red). The dashed lines stand for the calculated distributions based on the literature data and the continuous ones for the distribution fitted to the normalized, integrated experimental preedge intensity in the 8332−8340 eV energy range, represented by purple dots. The integrated preedge peak area value belonging to the solution with a 1:4 NiII/CN− ratio was chosen to be one, and all others were normalized to it [c(NiCl2) = 0.25 M, c(EDTA) = 0.30 M, and pH 11]. (D) Calculated XANES spectral series for the measured samples based on the determined nickel distribution (cf. panel C). The inset shows the calculated preedge region on a Lorentzian curve.

of the assumed absorbing species. Assuming one absorbing species (that is, only the tetracyano complex contributes to the preedge peak), the log K4 value is −0.84 ± 0.05 for the Ni(CN)53− complex, resulting in K4 = 0.14 ± 0.02, while these values are −0.60 ± 0.06 and 0.25 ± 0.04 for two absorbing species (accounting for the tetracyano and pentacyano species as well), respectively. The real stability constant has to be between these two values, which is in good agreement with the so-far-published values ranging from 0.168 to 0.4 (see the review work of Beck3 and references cited therein), as seen in Figure 3C. On the basis of this distribution diagram, the XANES and preedge spectral series were also calculated for the present experimental concentration series and displayed in Figures 3D. When the experimental and theoretical spectra in parts B and D of Figure 3, respectively, are compared, a very good qualitative agreement can be seen for the full XANES energy range. All spectral changes both before and after the K edge are very well reproduced, fortifying the preedge-based species identification and making it possible to reconstruct the XANES spectrum of the intermediate products from the measured CN− concentration-dependent data. In order to obtain these spectra, XANES of pure [NiEDTA]2− (measured at 0 cyanide concentration) and [Ni(CN)4]2− [the case of c(CN) = 1 M], weighed with their distribution-derived concentrations, were subtracted from the spectra with intermediate cyanide concentrations. Figure 4A shows the reconstructed spectrum for the low c(CN−) intermediate and the FEFF9-calculated [NiEDTA(CN)]3− XANES, comparing them to the [NiEDTA]2− spectra. Although the edge shift is not reproduced correctly, all other features of the XANES region of the reconstructed spectrum match very well with the calculations. This demonstrates the power of calculation-supported XAFS

which the edge shift is significantly lower. This property is met by the suggested ternary [NiEDTA(CN)]3−, adding more evidence for its occurrence at intermediate cyanide concentrations. On the basis of these findings and the unclarity of the literature about the composition of the intermediate species, several different models were tested to fit the experimental data with PSEQUAD.21 The average residual between the fitted and measured values was 0.023 absorbance unit (a.u.) for the bestfitting model, and the distribution diagram simulated based on this model is presented on Figure 3C with solid lines. This model contains four main components, namely, [NiEDTA(CN)]3−, [NiEDTA]2−, [Ni(CN)4]2−, and [Ni(CN)5]3−. (The residual values, as well as the stability constants obtained from the unsuccessful models, are discussed in the Supporting Information.) Our theoretical calculations (see Figure 3A) suggested that in the above-mentioned energy range the molar absorbance at the preedge XANES peak of the pentacyano complex is about onethird that of the tetracyano complex. However, this ratio is supposed to be even smaller based on literature findings;40 therefore, computations were carried out either supposing the 1:3 ratio for the molar absorbance of the pentacyano complex compared to the tetracyano or assuming that the tetracyano complex is the only absorbing species (the only species contributing to the absorbance at the preedge region). The log K2 of the [NiEDTA(CN)]3− complex was found to be 3.65 ± 0.14, which is in complete agreement with the value of 3.6 ± 0.2 found by Margerum et al.1 for the one cyano-coordinated NiEDTA molecule, giving an independent experimental confirmation that the intermediate product is [NiEDTA(CN)]3−. Therefore, the equilibrium constant of reaction (2) is (4.4 ± 1.4) × 103. This value is independent of the number E

DOI: 10.1021/acs.inorgchem.7b02311 Inorg. Chem. XXXX, XXX, XXX−XXX

Article

Inorganic Chemistry

used on a daily basis to determine both the stability constant and structure of the complex(es) in solution from the exact same sample. This method can be used effectively to study the reactions in which a geometric rearrangement takes place around the metallic center or its spin and/or oxidation state changes. Because these measurements are element-selective, the information comes only from the relevant part of the complex; thus, it cannot be hindered by signals from other parts of the ligand or other presenting metal ions, contaminations. Further insight can be gained with the advancement of this novel experimental tool. With improvement of the efficiency and spectral range, EXAFS and valence-to-core X-ray emission spectroscopy (vtc-XES) are foreseen to provide a more detailed description on the local atomic bond lengths and the chemical changes around the Ni ion, respectively.

■

ASSOCIATED CONTENT

S Supporting Information *

The Supporting Information is available free of charge on the ACS Publications website at DOI: 10.1021/acs.inorgchem.7b02311. Atomic coordinates of the BP86-optimized structures, a comparison of the calculated and experimental Ni K-edge positions, and a discussion of the unsuccessful models used to fit the experimental data (PDF)

■

AUTHOR INFORMATION

Corresponding Author

*E-mail: [email protected].

Figure 4. (A) Comparison of the calculated and either the experimental or reconstructed XANES spectra for the [NiEDTA]2− and [NiEDTA(CN)]3− complexes, respectively. (B) Comparison of the calculated and either the experimental or reconstructed XANES spectra for the [Ni(CN)4]2− and [Ni(CN)5]3− complexes, respectively. The detailed method of reconstruction is described in the text.

ORCID

Éva G. Bajnóczi: 0000-0002-8469-5887 György Vankó: 0000-0002-3095-6551 Notes

The authors declare no competing financial interest.

■

measurements: the full energy range with many structuresensitive features can be used as a fingerprint to separate different compounds, while the theoretical spectra help to identify them. The same method was used to confirm the [Ni(CN)5]3− molecular ion: panel B of Figure 4 shows a comparison of the calculated and reconstructed XANES of this compound, contrasted with the spectrum of the tetracyanonickel complex. Here again, the reconstructed spectrum mimics the changes predicted by theory.

ACKNOWLEDGMENTS This project was supported by the “Lendület” (Momentum) Program of the Hungarian Academy of Sciences (Grant LP2013-59), the European Research Council (Grant ERCStG-259709), the National Research, Development and Innovation Fund (NKFIH FK 124460), and the Government of Hungary and the European Regional Development Fund under Grant VEKOP-2.3.2-16-2017-00015. Z.N. acknowledges support from the Bolyai Fellowship of the Hungarian Academy of Sciences. The authors thank Csilla Bogdán, Gábor Endrő czi, and Gábor Peintler for their help in the experiments and their evaluation.

■

CONCLUSIONS AND PERSPECTIVES The paper follows the species evolution in the NiII−EDTA− CN− ternary system, in equilibrium with the varying CN− concentration. The transformations from the initial octahedral, paramagnetic [NiEDTA]2− complex through the ternary intermediate [NiEDTA(CN)]3− and the stable and wellknown square-planar, diamagnetic [Ni(CN) 4]2− to the pyramidal [Ni(CN)5]3− are traced by a novel von Hámos type laboratory XAFS spectrometer. This technique enabled not only unambiguous identification of the intermediate species with unleashing of the rich information from the preedge, edge position, and postedge regions of the XANES spectra but also determination of the corresponding stability constants with high precision. The experimental data were contrasted to modern quantum-chemical calculations, confirming the relationship between the structure and spectral changes. These results demonstrate that laboratory XAFS spectroscopy can be a versatile tool in coordination chemistry and can be

■

REFERENCES

(1) Margerum, D. W.; Bydalek, T. J.; Bishop, J. J. Kinetics of Nickel(II) Ligand Exchange Reactions: Cyanide Ion and (Ethylenedinitrilo)-tetraacetate Ion. J. Am. Chem. Soc. 1961, 83, 1791−1795. (2) Mccullough, R. L.; Jones, L. H.; Penneman, R. A. Infra-red and Visible Absorption Studies of the Pentacyanonickelate Ion in Aqueous Solution. J. Inorg. Nucl. Chem. 1960, 13, 286−297. (3) Beck, M. T. Critical Survey of Stability Constants of Cyano Complexes. Pure Appl. Chem. 1987, 59, 1703−1720. (4) Jones, J. P.; Margerum, D. W. Kinetics of the reactions between 1,2-diaminocyclohexanetetraacetatocobaltate(II) ion and cyanide ion. Inorg. Chem. 1969, 8, 1486−1490. (5) Coombs, L. C.; Margerum, D. W.; Nigam, P. C. Mechanism of Tetracyanonickelate Formation from Mono (aminocarboxylate) nickel (II) Complexes. Inorg. Chem. 1970, 9, 2081−2087.

F

DOI: 10.1021/acs.inorgchem.7b02311 Inorg. Chem. XXXX, XXX, XXX−XXX

Article

Inorganic Chemistry (6) X-Ray Absorption Principles, Applications, Techniques of EXAFS, SEXAFS and XANES Chemical Analysis: A Series of Monographs on Analytical Chemistry and Its Applications; Koningsberger, D., Prins, R., Eds.; John Wiley & Sons Ltd., 1987. (7) Stern, E. A.; Heald, S. M. In Handbook of Synchrotron Radiation: Basic Principles and Applications of EXAFS; Koch, E. E., Ed.; NorthHolland, 1983; Chapter 10; pp 995−1014. (8) Inada, Y.; Funahashi, S. Equilibrium and Structural Study of Chloro Complexes of Iron(III) Ion in Acidic Aqueous Solution by Means of X-Ray Absorption Spectroscopy. Z. Naturforsch., B: J. Chem. Sci. 1999, 54, 1517−1523. (9) Jullien, A.-S.; Gateau, C.; Kieffer, I.; Testemale, D.; Delangle, P. X-ray Absorption Spectroscopy Proves the Trigonal-Planar SulfurOnly Coordination of Copper(I) with High-Affinity Tripodal Pseudopeptides. Inorg. Chem. 2013, 52, 9954−9961. (10) Seidler, G. T.; Mortensen, D. R.; Remesnik, A. J.; Pacold, J. I.; Ball, N. A.; Barry, N.; Styczinski, M.; Hoidn, O. R. A laboratory-based hard x-ray monochromator for high-resolution x-ray emission spectroscopy and x-ray absorption near edge structure measurements. Rev. Sci. Instrum. 2014, 85, 113906. (11) Uhlig, J.; et al. Table-Top Ultrafast X-Ray Microcalorimeter Spectrometry for Molecular Structure. Phys. Rev. Lett. 2013, 110, 138302. (12) Joe, Y. I.; O’Neil, G. C.; Miaja-Avila, L.; Fowler, J. W.; Jimenez, R.; Silverman, K. L.; Swetz, D. S.; Ullom, J. N. Observation of iron spin-states using tabletop x-ray emission spectroscopy and microcalorimeter sensors. J. Phys. B: At., Mol. Opt. Phys. 2016, 49, 024003. (13) Schlesiger, C.; Anklamm, L.; Stiel, H.; Malzer, W.; Kanngießer, B. XAFS spectroscopy by an X-ray tube based spectrometer using a novel type of HOPG mosaic crystal and optimized image processing. J. Anal. At. Spectrom. 2015, 30, 1080−1085. (14) Németh, Z.; Szlachetko, J.; Bajnóczi, É.; Vankó, G. Laboratory von Hámos X-ray spectroscopy for routine sample characterization. Rev. Sci. Instrum. 2016, 87, 103105. (15) Szlachetko, J.; et al. A von Hamos x-ray spectrometer based on a segmented-type diffraction crystal for single-shot x-ray emission spectroscopy and time-resolved resonant inelastic x-ray scattering studies. Rev. Sci. Instrum. 2012, 83, 103105. (16) Ward, L. G. L. In Inorganic Syntheses; Cotton, F. A., Ed.; McGraw-Hill, Inc., 1972; Vol. XIII; Chapter 30, pp 154−164. (17) Sipos, P. M.; Hefter, G.; May, P. M. Viscosities and Densities of Highly Concentrated Aqueous MOH Solutions (M+ = Na+, K+, Li+, Cs+, (CH3)4N+) at 25.0°C. J. Chem. Eng. Data 2000, 45, 613−617. (18) Sipos, P.; May, P. M.; Hefter, G. T. Carbonate removal from concentrated hydroxide solutions. Analyst 2000, 125, 955−958. (19) Ravel, B.; Newville, M. ATHENA, ARTEMIS, HEPHAESTUS: data analysis for X-ray absorption spectroscopy using IFEFFIT. J. Synchrotron Radiat. 2005, 12, 537−541. (20) Savitzky, A.; Golay, M. J. E. Smoothing and Differentiation of Data by Simplified Least Squares Procedures. Anal. Chem. 1964, 36, 1627−1639. (21) Zékány, L.; Nagypál, I.; Peintler, G. PSEQUAD for Chemical Equilibria. Update 5-5.10, 2000−2008; 1987−2012); http://www.staff. u-szeged.hu/peintler/download/psequad/psequad.zip, last accessed on October 17, 2017. (22) Neese, F. The ORCA program system. Wiley Interdiscip. Rev.: Comput. Mol. Sci. 2012, 2, 73−78. (23) Schäfer, A.; Horn, H.; Ahlrichs, R. Fully optimized contracted Gaussian basis sets for atoms Li to Kr. J. Chem. Phys. 1992, 97, 2571− 2577. (24) Schäfer, A.; Huber, C.; Ahlrichs, R. Fully optimized contracted Gaussian basis sets of triple zeta valence quality for atoms Li to Kr. J. Chem. Phys. 1994, 100, 5829−5835. (25) Klamt, A.; Schuurmann, G. COSMO: a new approach to dielectric screening in solvents with explicit expressions for the screening energy and its gradient. J. Chem. Soc., Perkin Trans. 2 1993, 0, 799−805.

(26) Becke, A. D. Density-functional exchange-energy approximation with correct asymptotic behavior. Phys. Rev. A: At., Mol., Opt. Phys. 1988, 38, 3098−3100. (27) Perdew, J. P. Density-functional approximation for the correlation energy of the inhomogeneous electron gas. Phys. Rev. B: Condens. Matter Mater. Phys. 1986, 33, 8822−8824. (28) Becke, A. D. Density-functional thermochemistry. III. The role of exact exchange. J. Chem. Phys. 1993, 98, 5648−5652. (29) Lee, C.; Yang, W.; Parr, R. G. Development of the Colle-Salvetti correlation-energy formula into a functional of the electron density. Phys. Rev. B: Condens. Matter Mater. Phys. 1988, 37, 785−789. (30) Stephens, P. J.; Devlin, F. J.; Chabalowski, C. F.; Frisch, M. J. Ab Initio Calculation of Vibrational Absorption and Circular Dichroism Spectra Using Density Functional Force Fields. J. Phys. Chem. 1994, 98, 11623−11627. (31) Reiher, M.; Salomon, O.; Artur Hess, B. Reparametrization of hybrid functionals based on energy differences of satates of different multiplicities. Theor. Chem. Acc. 2001, 107, 48−55. (32) Tao, J.; Perdew, J. P.; Staroverov, V. N.; Scuseria, G. E. Climbing the Density Functional Ladder: Nonempirical Meta-Generalized Gradient Approximation Designed for Molecules and Solids. Phys. Rev. Lett. 2003, 91, 146401. (33) Perdew, J. P.; Tao, J.; Staroverov, V. N.; Scuseria, G. E. Metageneralized gradient approximation: Explanation of a realistic nonempirical density functional. J. Chem. Phys. 2004, 120, 6898−6911. (34) Rehr, J.; Kas, J. J.; Vila, F. D.; Prange, M. P.; Jorissen, K. Parameter-free calculations of X-ray spectra with FEFF9. Phys. Chem. Chem. Phys. 2010, 12, 5503−5513. (35) Neese, F.; Olbrich, G. Efficient use of the resolution of the identity approximation in time-dependent density functional calculations with hybrid density functionals. Chem. Phys. Lett. 2002, 362, 170−178. (36) Petrenko, T.; Kossmann, S.; Neese, F. Efficient time-dependent density functional theory approximations for hybrid density functionals: Analytical gradients and parallelization. J. Chem. Phys. 2011, 134, 054116. (37) DeBeer George, S.; Petrenko, T.; Neese, F. Prediction of Iron K-Edge Absorption Spectra Using Time-Dependent Density Functional Theory. J. Phys. Chem. A 2008, 112, 12936−12943. (38) Chandrasekaran, P.; Stieber, S. C. E.; Collins, T. J.; Que, L., Jr.; Neese, F.; DeBeer, S. Prediction of high-valent iron K-edge absorption spectra by time-dependent Density Functional Theory. Dalton Trans. 2011, 40, 11070−11079. (39) Galoisy, L.; Calas, G. XANES and Crystal Field Spectroscopy of Five-coordinated Nickel(II) in Potassium-Nickel Phosphate. Mater. Res. Bull. 1993, 28, 221−228. (40) Furenlid, L. R.; Renner, W.; Fujita, E. XAS studies of Ni (I), Ni (II), and Ni (III) complexes. Phys. B 1995, 208−209, 739−742.

G

DOI: 10.1021/acs.inorgchem.7b02311 Inorg. Chem. XXXX, XXX, XXX−XXX