Gas Phase Ligand Exchange

Article pubs.acs.org/IC

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

Unexpected Sequential NH3/H2O Solid/Gas Phase Ligand Exchange and Quasi-Intramolecular Self-Protonation Yield [NH4Cu(OH)MoO4], a Photocatalyst Misidentified before as (NH4)2Cu(MoO4)2

Inorg. Chem. Downloaded from pubs.acs.org by UNIV OF LOUISIANA AT LAFAYETTE on 10/30/18. For personal use only.

István E. Sajó,† László P. Bakos,‡ Imre M. Szilágyi,‡ György Lendvay,§ József Magyari,∥ Miklós Mohai,§ Á gnes Szegedi,§ Attila Farkas,⊥ Anna Jánosity,§ Szilvia Klébert,§ and László Kótai*,§,# †

University of Pécs, János Szentágothai Research Centre, Pécs, H-7624, Hungary Budapest University of Technology and Economics, Department of Inorganic and Analytical Chemistry, Müegyetem rakpart 3, Budapest, H-1111, Hungary § Research Centre for Natural Sciences, Hungarian Academy of Sciences, Magyar Tudósok krt. 2, Budapest, H-1519, Hungary ∥ Department of Chemistry, Biochemistry and Environmental Protection, Faculty of Sciences, University of Novi Sad, Trg Dositeja Obradovića 3, Novi Sad, 21000, Serbia ⊥ Budapest University of Technology and Economics, Department of Organic Chemistry, Müegyetem rakpart 3, Budapest, H-1111, Hungary # Deuton-X Ltd., H-2030, É rd, Selmeci u. 89, Hungary ‡

S Supporting Information *

ABSTRACT: [NH4Cu(OH)MoO4] as active photocatalyst in the decomposition of Congo Red when irradiated by UV or visible light has been prepared in an unusual ammonia/water ligand exchange reaction of [tetraamminecopper(II)] molybdate, [Cu(NH3)4]MoO4. [Cu(NH3)4]MoO4 was subjected to moisture of open air at room temperature. Light blue orthorhombic [Cu(NH3)(H2O)3]MoO4 was formed in 2 days as a result of an unexpected solid/gas phase ammonia−water ligand exchange reaction. This complex does not lose its last ammonia ligand on further standing in open air; however, a slow quasiintramolecular (self)-protonation reaction takes place in 2−4 weeks, producing a yellowish-green microcrystalline material, which has been identified as a new compound, [NH4Cu(OH)MoO4], (a = 10,5306 Å, b = 6.0871 Å, c = 8.0148 Å, β = 64,153°, C2, Z = 4). Mechanisms are proposed for both the sequential ligand exchange and the self-protonation reactions supported by ab initio quantum-chemical calculations and deuteration experiments as well. The [Cu(NH3)(H2O)3]MoO4 intermediate transforms into NH4Cu(OH)(H2O)2MoO4, which loses two waters and yields [NH4Cu(OH)MoO4]. Upon heating, both [Cu(NH3)4]MoO4 and [Cu(NH3)(H2O)3]MoO4 decompose, losing three NH3 and three H2O ligands, respectively, and stable [Cu(NH3)MoO4] is formed from both. The latter can partially be hydrated in boiling water into [NH4Cu(OH)MoO4. This compound can also be prepared in pure form by boiling the saturated aqueous solution of [Cu(NH3)4]MoO4. All properties of [NH4Cu(OH)MoO4] match those of the active photocatalyst described earlier in the literature under the formulas (NH4)2[Cu(MoO4)2] and (NH4)2Cu4(NH3)3Mo5O20.

■

materials such as catalysts,2a,b disinfectants,3 or electrode materials.4a−c Complexes of copper(II) molybdate containing ammonia and/or ammonium ion have been found to manifest

INTRODUCTION Tetraoxometallates of tetraamminecopper(II) cation have been known for more than a century.1a,b They are generally stable compounds; their crystals are blue as typical to copper− ammonia complexes. They are important precursors or intermediates in the synthesis of many industrially important © XXXX American Chemical Society

Received: August 9, 2018

A

DOI: 10.1021/acs.inorgchem.8b02261 Inorg. Chem. XXXX, XXX, XXX−XXX

Article

Inorganic Chemistry

4:5 in 3, and 1:1 in 4, which is hard to reconcile with the same structure suggested by the XRD results. The obvious contradictions concerning the structure and composition of compounds with important industrial applications require clarification. In order to resolve these controversies concerning the ammonia and ammonium-ion containing derivatives of copper molybdate, as a part of our continuing investigations on tetraoxometallates,11a−e we synthesized them in the known and also in some new reaction routes and determined the composition and properties of the intermediate and final products. The following compounds and intermediates are described (Table 1).

some unexpected structural and synthetic complications. Such is [Cu(NH3)4]MoO4 (1), for which two different appearances and XRD patterns were reported.5,6 The other concerns the series 2−4, namely (NH4)2[Cu(MoO4)2], (2), (a well-known photocatalyst,2a,b,7), (NH4)2Cu4(NH3)3Mo5O20 (3), and [NH4Cu(OH)MoO4] (4): the same XRD pattern has been assigned to all three in refs 8, 9a, and b and in the present work. This paper intends to show that the two phenomena are strongly connected and to resolve the ambiguities in the properties of [Cu(NH3)4]MoO4 (1) and in the composition of three materials producing the same XRD pattern. We also report on the investigation of an unexpected sequential solidstate reaction of [Cu(NH3)4]MoO4 with H2O vapor, going through Intermediates [Cu(H2O)4−n(NH3)n]MoO4 (n = 1−3) (5−7), respectively), leading to the formation of [NH4Cu(OH)MoO4] (4) as the final product. The compounds involved in our studies have attracted considerable interest recently. Three synthesis routes for tetraamminecopper(II) molybdate, [Cu(NH3)4]MoO4 (1) have been reported in the literature and identified and characterized the products. Route 15 reacted basic copper(II) molybdate and aqueous ammonia, precipitating the product by ethanol addition. We denote the product of this route by 1a. It was described as an unstable blue compound, which decomposed easily with ammonia release. Being unstable, the crystal structure of the product could not be determined, but the presence of [Cu(NH3)4]2+ and MoO42− ions was detected by vibrational spectroscopy, and elemental analysis agreed with the formula [Cu(NH3)4]MoO4. Route 26 prepared CuMoO4 from CuO and MoO3 in a 1:1 molar ratio by heating at 630 °C for 2 h and reacted it with aqueous ammonia. Crystallization of the cold (around 5 °C) solution, without adding ethanol, led to dark blue blocks of their product, denoted here as 1b. When studying its thermal properties, they observed that it decomposes at relatively high temperature (Tdec > 125 °C). Route 310 followed the synthetic path of Mueller and isolated a microcrystalline powder (called here 1c) by adding dioxane to induce crystallization. The reported IR spectra of compounds 1a−c are identical,5,6,10 suggesting that their composition is the same. On the other hand, the powder XRD patterns reported for 1a5 and 1b6 are not the same, and the reported appearance and stability of the products of the three routes are definitely different, which is an issue to be resolved. Concerning the triplet with the same XRD pattern, the compound described as (NH4)2[Cu(MoO4)2] (2) has been used in different areas such as photocatalysts in the degradation of dyes,2a,b in prevention of healthcare-associated infections3 or as a precursor in the preparation of Cu3Mo2O9, for use as a nanoplate anode material in high storage-capacity lithium-ion batteries.4a−c There are several synthetic routes reported for this compound,5−8 but not only its structure but even its identity seem to be uncertain. For example, the same XRD pattern has been assigned to another compound, the complex reported to be (NH4)2Cu4(NH3)3Mo5O20 (3),9a,b but the cells were identified to be monoclinic and triclinic for 2 and 3, respectively.8,9a,b Furthermore, in the studies reported in the present paper on the investigation of the instability of 1 observed by Mueller et al.,5 1 was found to react with the humidity of open air and produce [NH4Cu(OH)MoO4] 4, whose XRD pattern proved to be identical to those reported for both 2 and 3. Note that the Cu:Mo atom ratio is 1:2 in 2,

Table 1. Compounds and Intermediates Studied and Described Compounds Label

Previously reported

Present work

1a−c 2 3 4 5 6 7 8

[Cu(NH3)4]MoO4 (NH4)2[Cu(MoO4)2] (NH4)2Cu4(NH3)3Mo5O20 Unknown Unknown Unknown Unknown Unknown

[Cu(NH3)4]MoO4 [NH4Cu(OH)MoO4] [NH4Cu(OH)MoO4] [NH4Cu(OH)MoO4] [Cu(H2O)3(NH3)]MoO4 [Cu(H2O)2(NH3)2]MoO4 [Cu(H2O)(NH3)3]MoO4 [Cu(NH3)MoO4]

In the reaction of 1 with H2O, intermediates [Cu(H2O)4−n(NH3)n]MoO4 with n = 1−3 (5−7) as well as the final product 4 have been identified. The blue intermediate compound [Cu(H2O)3(NH3)]MoO4 (5) as a product of an unusual solid phase ammonia/water ligand exchange reaction was identified, and its role in the formation of compound 4 from 1 has been elucidated. Compound 4, formed from 1, has been thoroughly characterized and was proved to be identical to the compound that was widely studied under the belief it was 2. In the rest of this paper, we first present the synthesis of [Cu(NH3)4]MoO4, the explanation for the different XRD patterns reported in the literature and for the sensitivity of the microcrystalline form of 1 to humidity, followed by the identification of the final product, [NH4Cu(OH)MoO4] 4, and its characterization. Then, we describe the mechanism of the NH 3 to H 2 O ligand exchange in the series [Cu(H2O)4−n(NH3)n]MoO4 with n going from 4 to 1. Finally, we demonstrate that the thermal decomposition of compounds 4 and 5 allows us to synthesize the coordinatively unsaturated compound 8.

■

RESULTS AND DISCUSSION Synthesis of [Cu(NH3)4MoO4] (1) and Its Transformation to [NH4Cu(OH)MoO4] (4). The differences in the properties and XRD patterns of nominally the same compound (1a−c) prepared in different ways could be understood by considering them as polymorphs. Our DSC studies, however, did not show signs of any phase transition, excluding the possibility of polymorphism. Therefore, the described reaction routes and isolation methods have been tested to find the reasons for differences in the characteristics of the samples. As a first step, we repeated the synthesis routes R1 and R2 (R3 being just a modification of route R1). The major differences between routes R1 and R2 are (1) the identity of B


Article

Inorganic Chemistry educts and (2) the method used to initiate crystallization. In order to understand whether the former or the latter is responsible for the difference in the properties of the product, we combined the educts of both routes with both crystallizers. First we tested the role of the educts when ethanol was used to induce precipitation as in R1. When basic (as in R1) or regular (as in R2) copper(II) molybdate was dissolved in concentrated aqueous NH3 solution, deep blue reaction mixtures were formed. Then both solutions were mixed with a large excess of ethanol when purplish blue microcrystalline precipitate formed in either case. The products made from the educts of either R1 or R2 easily lost ammonia and could only be dried in ammonia atmosphere (generated in situ from NH4Cl) in a desiccator containing an alkaline drying agent (CaO).25 The XRD of both primary products were found to be identical, irrespective of the nature of the educts, and were very similar to that reported by Mueller.5 The dried precipitates made via either route proved to be stable; they could be stored without decomposition in a dark bottle for weeks. In contrast, when the freshly made samples were left wet or when dry samples were kept in humid air, both decomposed quickly, releasing ammonia exactly as it was described by Mueller et al.5 Based on the observation that the properties of the precipitates were identical when ethanol was used to initiate crystallization from the reaction mixtures of either route, we concluded that the same compound was formed in both cases, which is the same as 1a observed by Mueller et al.5 Formation of a decomposition product was observed when the XRD was taken after half an hour or later. The disappearance of the primary product was completed in 1−2 days under ambient conditions. When the deep blue solutions made as in the first steps of routes R1 or R2 were left to stand in a refrigerator around 0 °C (as in R2), dark blue crystals of the compound we call 1b were formed in both cases. The XRD of these two products and the two dry powders formed with ethanol addition were found to be identical. The elemental analysis of the dry powder (1a) and crystalline (1b) samples confirmed that they are chemically identical: the composition of both is [Cu(NH3)4]MoO4.26 The only difference between the two is that no large crystals were allowed to grow by the fast nucleation induced by ethanol addition, while slow crystallization in a cold solution favored the formation of large crystals. Evaluation of Reasons for Contradictions in Identification of [Cu(NH3)4MoO4] (1). We conjecture that the reason why different XRD patterns were found in ref 5 vs ref 6 for the same compound is that the microcrystalline form of the product, 1a, “ages” quickly, and if the diffractogram was taken a few hours after synthesis, the sample was already not pure [Cu(NH3)4]MoO4. This is supported by the fact that Mueller et al. concluded that proper indexing of the structure is not possible (because of the presence of contaminants already formed in the samples). The large crystals of 1b seem to be much more stable in the open air than the microcrystalline samples of 1a. It took an entire day until a light blue layer of decomposition product appeared on the surface of 1b crystals. This means that the composition of 1a and 1b is identical, and the only difference between them is the crystal size. The light blue decomposition product, however, slowly transforms further into a yellowish green material. When analyzing the final product of the transformation of 1a, we found it to be [NH4Cu(OH)MoO4] (4):

[Cu(NH3)4 ]MoO4 + H 2O = [NH4Cu(OH)MoO4 ] + 3NH3

The reaction can be accelerated by boiling the aqueous solution of compound 1 for 10 min. Under these conditions, one would expect the formation of (NH4)2MoO4 and Cu(OH)2.10 The removal of ammonia by boiling not only eliminates a part of the ammonia molecules from the complex equilibrium12a−c but also decreases the pH and increases the amount of ammonium ions, which leads to deposition of the sparingly soluble [NH4Cu(OH)MoO4] (4) instead of the expected Cu(OH)2. The XRD of this product proved to be identical to the one reported by Astier et al.9a,b for what they considered to be (NH4)2Cu4(NH3)3Mo5O20 (3) and to that presented by Garin and Costamagna8 for the compound they identified as (NH4)2[Cu(MoO4)2] (2). Based on our results on the synthesis of [Cu(NH3)4]MoO4 (1) and on its properties, our four main conclusions are (1) The product of both synthesis routes proposed in the literature is the expected compound, 1. (2) The only difference is that fast nucleation yields a microcrystalline, slow nucleation macroscopic crystalline appearance of [Cu(NH3)4]MoO4(1);. (3) [Cu(NH 3 ) 4 ]MoO 4 (1) is sensitive to humidity, especially in the microcrystalline form having high specific surface area, and reacts with H2O to yield [NH4Cu(OH)MoO4] (4). (4) [NH4Cu(OH)MoO4] (4) has the same XRD pattern as reported for an important photocatalyst (2). From these conclusions two questions arise: (a) how can three different compounds yield the same XRD pattern and (b) why is [Cu(NH3)4]MoO4 sensitive to H2O and how can [NH4Cu(OH)MoO4] be formed from it. In the following, we address the first question by a detailed characterization of [NH4Cu(OH)MoO4] (4), after which we illuminate the sequence of steps in the reaction of [Cu(NH3)4MoO4] (1) when H2O is present in the atmosphere around it. [NH4Cu(OH)MoO4] (4) Is the Compound Which Yielded the XRD Pattern Reported Earlier for Other Compounds (2 and 3). First we repeated the synthesis route reported by Garin and Costamagna8,14 to yield (NH4)2[Cu(MoO4)2], (2) as well as its modifications used by Xue7 and refs 2a and b and that applied by Astier9 to synthesize (NH4)2Cu4(NH3)3Mo5O20 (3). Each of these routes was found to yield the same product, characterized in detail below. The elemental analysis proved this compound to correspond to the formula [NH4Cu(OH)MoO4], 4. This means that the results of the investigations of the nanochemistry of what was thought to be 2 actually apply not to 2 but to 4 instead. Physicochemical Properties of [NH4Cu(OH)MoO4] (4). Compound 4 is a yellowish green powder (110 μm platelets, Figure 1), stable in air and in boiling water for a long time. It is insoluble in water and organic solvents, which suggests this compound is a polymer. On the other hand, it dissolves easily in excess of concentrated ammonia with dark blue color. From this solution, 1 can be crystallized (with all properties identical to those described above), in contrast to the composition (NH4)2[Cu(NH3)2(MoO4)] proposed by Garin et al.8 A related compound, [NaCu(OH)MoO4] has already been synthesized.15 We have attempted to exchange the ammonium C


Article

Inorganic Chemistry

Figure 1. SEM image of [NH4Cu(OH)MoO4] (4).

ion to Na+ in [NH4Cu(OH)MoO4] with NaOH, NaHCO3, and NaCl according to the reaction: [NH4Cu(OH)MoO4 ] + NaX = NH4X + Na[Cu(OH)MoO4 ]X = OH, HCO3 or Cl

Treatment of 4 with 0.5 M sodium hydroxide solution at room temperature resulted in a light blue Cu(OH)2 precipitate (blackened on boiling, Cu(OH)2 → CuO). There was no reaction with saturated aq. NaHCO3 at room temperature, but when the solution was boiled, 4 decomposed as occurred in the case of sodium hydroxide. The attempt to do a simple ion exchange with saturated aqueous sodium chloride solution did not result in any changes in compound 4 either. This means that either the solubility of 4 is extremely small or the NH4+ ion is not in a changeable position in the crystal lattice. Thermal Studies on [NH4Cu(OH)MoO4] (4). Thermal decomposition of [NH4Cu(OH)MoO4] has been studied both in He and in air atmospheres between 25 and 800 °C (heating rate 5°/min). The measured mass loss in an inert atmosphere was found to be 13.2% (Table 2), close to the calculated 13.5% Table 2. Thermal Decomposition Characteristics of Compound 4 at 5 °C/min Heating Rate in Inert Atmosphere as Well as in Air Peak temp., °C Stage

Temp. range, °C

1 2

250−310

1 2

250−320

DTG

DSC

Mass loss, %

Inert Atmosphere (He) 270 290 13.2 278 292

294 310

Under Air 297 11.2 312

TG-MS fragments NH3, H2O, NO NH3, H2O, N2O, NO NO, N2O, NH3, H2O

Figure 2. TG-DSC (a) and TG-MS spectra (m/z = 15, 16, 17, and 18) (b) and (m/z = 30, 44) (c) of [NH4Cu(OH)MoO4] (4) under inert atmosphere.

for the release of 1 mol of NH3 and H2O each from [NH4Cu(OH)MoO4]. According to the evolved gas analysis, the main gaseous decomposition products were water and ammonia. Some minor amounts of nitrous gases like NO or N2O were also detected (Figure 2a−c).

In the presence of oxygen, the decomposition starts in the same way as under inert atmosphere (Supporting Information Figure S1), but before completing the ammonia release, the residual coordinated ammonia is oxidized with the formation of oxygen-containing compounds at higher temperatures D


Article

Inorganic Chemistry causing an unexpected mass increase of the sample. The results are summarized in Table 2 and Supporting Information Figure S1. The relative intensities of the NH3:NH2:NH:N and H2O:OH:O fragments (Table 3)16 show that the m/z = 17 Table 3. Distinctive Relative Intensities of m/z Values of Analyzed Gases According to Databasesa m/z

N2

NH3

14 15 16 17 18 30 44

7.2%

2.2% 7.5% 80.1% 100%

H2O

1.1% 23.0% 100%

N 2O

NO

NO2

12.9%

7.5% 2.4% 1.5%

9.6% 22.3%

100%

100%

5.0%

31.1% 100%

a

Mass spectrometer Netzsch QMS 403C and http://webbook.nist.gov/.

and 16 signals contain more than one component (the m/z = 17 signals may belong to OH and NH3 and m/z = 16 signals may belong to O and NH2). The signals at m/z = 15, however, belong only to the ammonia while the signal at m/z = 14 may be a fragment of ammonia or nitrogen oxides. Small amounts of nitrogen oxides (m/z = 44 and m/z = 30, N2O and NO, respectively) could also be found. The high intensity of the m/ z = 30 peak as compared with that of m/z = 44 shows (Table 3) that NO is not only the fragment of N2O but appears as individual oxidation product as well. In the redox reactions observed in inert atmosphere, the oxidant may be the molybdate and/or the copper(II) ion, resulting in the formation of the X-ray amorphous residues of oxygen-deficient copper molybdenum oxides. In the decomposition in air, the final product was found to be crystalline copper molybdate.17 The DSC curve confirms the multistep character of the decomposition process both in inert atmosphere and in air. The peak temperature under He is 292 °C, but two further heat effects could also be observed at 471 and 575 °C, both without mass loss. There are six known polymorphs of CuMoO417,18 and the DSC signal at 575 °C unambiguously belongs to the room-temperature (α) → high temperature (β) phase transition of the CuMoO4. There is no known phase transition between room temperature and 575 °C. Thus, the peak observed at 471 °C might only be attributed to the crystallization of the amorphous CuMoO4 formed from 4 during heating. Interrupting the thermal decomposition process at 350 °C in air led to a black X-ray amorphous product. On the other hand, longer heating (2 h) at 400 °C resulted in pure crystalline αCuMoO4. This shows that the decomposition in air is controlled not only by the temperature but by kinetical factors as well. XPS Characterization of [NH4Cu(OH)MoO4] (4). There is only minimal information about the XPS spectroscopy of compound 4. Pal et al.2a studied some of the XPS characteristics of what they thought to be (NH4)2Cu(MoO4)2 but was in fact [NH4Cu(OH)MoO4]. Our XPS results (Figure 3) unambiguously show that the molybdenum is in hexavalent (molybdate) state (signals corresponding to Mo 3d5/2 and Mo 3d3/2 were found at 232.2 and 235.6 eV, respectively).2a No nitrogen was detected in the 6 nm thick surface layer of the

Figure 3. XPS results of [NH4Cu(OH)MoO4] (4).

compound, and a part of copper(II) was found to be reduced to copper(I) (manifested in the Cu 2p3/2 peaks at 932.1 and 935.5 eV for Cu1+ and Cu2+, respectively, and the Cu 3p values found at 75.1 and 77.2 eV for Cu1+, as well as 77.3 and 79.2 eV for Cu2+). Some C 1s signals were also detected, but they belong to organic contaminants. The lack of nitrogen (expected to be present in the NH4+ ions) and the appearance of copper(I) in the surface layer can be explained by assuming that some redox reaction between the ammonium ion and the copper(II) center is induced by the X-ray irradiation during the measurements, yielding gaseous NE


Article

Inorganic Chemistry

the number of peaks were determined by band decomposition (Figure 4).

containing redox products (N2, N-oxides), which were removed by vacuum. In the following, we analyze the XPS spectra reported by Pal et al.2a for their sample that we have shown above to be compound 4, and we argue that they do not contradict the assignment we presented above. Inspection of the spectrum measured by Pal et al. reveals that the Cu peak at about 933.5 eV is wider than what is typical for copper(II), which may be caused by the appearance of the close-lying copper(I) peak, so that their spectrum does not exclude the presence of copper(I) in the surface layer. Further analysis of the spectrum of Pal et al. shows that it does not contradict our conclusion that N is absent in the surface layer of 4. In their spectrum analysis, the 402.4 eV peak was assigned as the N 1s signal of the N atom of the ammonium ion they expected to be on the surface of crystals of compound 2. We note, however, that the position of this band coincides with that of the Mo 3p3/2 peak. We also observed that when the spectra are taken using Mg Kα irradiation, the Cu L2M23M45 Auger signal also appears at 402 eV. Using Al Kα irradiation, the Cu Auger peak shifts and the intensity of the 402 eV peak decreases. This means that the peak at 402.4 eV, when taken with Mg Kα irradiation, is surely a combination of the Mo 3p3/2 emission and the Cu-Auger signal. In principle, the N 1s emission might also be a component of this signal. This, however, can be excluded because the concomitant Auger signals of nitrogen could be detected neither with using MgKα nor with AlKα excitation. This means that what Pal et al.2a assigned as N 1s is instead a combination of Mo and Cu peaks; thus, very probably no ammonium ion was present in the upper irradiated layer of their sample, either. Vibrational Spectroscopy of [NH4Cu(OH)MoO4] (4). The IR and Raman spectra of compound 4 (Supporting Information Figure S2 and S3, respectively) were recorded at room temperature. The assignment of the IR and Raman bands is given in Table 4. The IR spectrum is relatively complicated in the range of symmetric and asymmetric molybdate stretching bands (ν1 and ν3, respectively); thus,

Figure 4. Deconvolution of the IR spectrum of compound 4 in the range of the ν1 and ν3 Mo−O bands.

The isolated tetrahedral molybdate ion has four normal modes, ν1 (νs)(A1), ν2 (δs)(E), ν3 (νas)(F2), and ν4 (δas)(F2). The peak corresponding to the ν1 symmetric stretching mode can be unambiguously identified as the singlet at 921 cm−1 in the IR and the singlet with increasing intensity in the Raman spectrum at 905 cm−1. The Raman band of ν3(Mo−O) is split into three parts and the deconvolution of the IR spectrum (Figure 4) also shows the triplet (F2) nature of the band. According to the correlation table (Table 5), the molybdate has bidentate (C2v) (or lower symmetry) coordination mode.19a−g Table 5. Correlation Table of Tetrahedral Molybdate and Ammonium Ions19g ν1 ν2 ν3 ν4

Table 4. Assignment of IR and Raman Bands of [NH4Cu(OH)MoO4] (4)

Td

C3v

C2v

A1 E F2 F2

A1 E A1 + E A1 + E

A1 A1 + A2 A1 + B1 + B2 A1 + B1 + B2

Compound 4 IR

Raman

Assignment

505 618 696 761 819 884sh 921 1034 1288 1413 1602 2023 2807 3012 3235 3413 3478

481

νs(CuOMo), CuOH(bridge) ν(Cu−O−Mo) δ(CuOH) νas(Mo−O)

750 825 868 905

The chelating bidentate coordination mode (η2-MoO4) should result in a significant shift to higher frequencies in the highest frequency component of the ν3 band toward the C2v‑μ2-bridged or lower symmetry ones.19a−g This difference is Δν = 50 cm−1 for compound 1 (the A1 and B1 components are located at 884 and 834 cm−1, respectively) in the present case, which is a much lower shift than that observed in other isotypic tetrahedral η2-C2v anions (Δν ∼ 95−124 cm−1 for sulfates and chromates.19a−g) The multiplet structure of the Cu−O Raman bands in the far-IR region can be attributed to the presence of different Cu−O environments, the lattice symmetry effect, and the presence of hydrogen bonds as well. The bands at 476 or 480 cm−1 in the IR and Raman spectra, respectively, probably belong to symmetric Cu−O−Mo and Cu−OH-Cu vibrations. The deformation band at 1035 cm−1 in the IR spectrum characterizes δOH and/or δCu‑(OH)‑Cu vibrations. The band at 1288 cm−1 (close to the deformation

νs(Mo−O) δCu(OH)Cu(bridge), δ(OH) Cu−O−Cu bridge δas(NH4+) δs(NH4+) δs(NH) + ν6(lattice) 2δas(NH) νs(NH) νas(NH) νs(NH) + ν5(lattice) ν(OH) F


Article

Inorganic Chemistry band of the coordinated ammonia ligand Cu-NH3 band, but at a distinctly different frequency) can be assigned to the Cu− O(X)−Cu bridged structure, where X may be H or molybdate as well. The normal modes of the tetrahedral ammonium ion (Table 5) could also be found. The presence of a combination band of the symmetric deformation and a lattice vibration at 2023 cm−1 (ν2 + ν6) unambiguously show the presence of hindered rotation, i.e., that the ammonium is in an environment of strong hydrogen bonds.19a This agrees with the observation that the hydrogens connected to the N atom cannot be exchanged by deuterium (see later). Crystal Structure of [NH4Cu(OH)MoO4] (4). Lattice parameters of compound 4 according to the powder XRD pattern we obtained for [NH4Cu(OH)MoO4] at 25 °C have been compared with those reported in the literature for compounds 2 and 3. The patterns were found to be identical except for some differences in resolution. As we have shown above, the synthesis procedures used to produce the latter compounds, in fact, yielded microcrystals of [NH4Cu(OH)MoO4]. From the same XRD pattern for the same compound, three different sets of lattice parameters have been derived by the different groups (see Table 6, Supporting Information

Table 7. Apparent Reaction Rate Constants of the Photodegradation of Congo Red at 375 nm Irradiation (18 W, 0.03% catalyst-compound 4) Substrate

System Space group a, Å b, Å c, Å α, deg β, deg γ, deg V, Å3 Z Dcalcd, g/cm3 Dm, g/cm3

28

39a,b

Monoclinic C2 10.5306 6.0871 8.0148 90 64.153 90 462.36 4 3.71 3.69

Triclinic P1̅ 6.155 7.158 7.516 108.52 93.99 113.72 279.94 1 2.69 2.64

Monoclinic

R2

1.2 6.1

0.73244 0.92075

Congo red, 2.10 M, without catalyst Congo red, 2.10−5 M, with catalyst

Under the conditions of our experiments, compound 4 acts as a photocatalyst and speeds up the decomposition of Congo red by about a factor of 5. This observation supports those made by Pal et al. and shows that compound 4 is a promising photocatalyst in both the UV and visible light wavelength range. Mechanism of the Transformation of [Cu(NH3)4MoO4] (1) to [NH4Cu(OH)MoO4] (4) in Humid Air. (a) Formation of [Cu(H2O)3(NH3)]MoO4 (5). As mentioned earlier, in the presence of H2O vapor, [Cu(NH3)4]MoO4 (1) can be converted to [NH4Cu(OH)MoO4] (4). The process is spontaneous and relatively slow, but can be accelerated, and it is challenging to understand how it takes place. For this purpose, we have explored the possible steps of the 1 to 4 conversion when solid [Cu(NH3)4]MoO4 (1) is subjected to humid air. First, we note that compound 1 is stable for weeks in a closed crucible in the dry state or in vacuum but decomposes in hours in the presence of humidity: When left in open air, it is converted to a product, compound 5, identified as [Cu(H2O)3(NH3)]MoO4. The reaction is completed in 2 days. The elemental analysis, TG-MS, and diffuse-reflectance visible spectroscopic results all confirm the formula with three water ligands and one ammonia ligand. The absorption maximum of compound 1 observed at 583 nm is shifted to 752 nm in compound 5. The former maximum is known to be characteristic to the tetraamminecopper(II) ion (CuN4 chromophore), while the latter corresponds to the [(ammine)(triaquo)copper(II)] cation (CuO3N chromophore) (Table 8).20a−d,21

Table 6. Powder XRD Data (25 °C) of Compounds 2−4 4

kapparent [10−4 min−1]

−5

5.134 3.051 7.232 90 103.8 90 110.01 1 17.22

Table 8. UV−Vis Characteristics of the [Cu(H2O)4−n(NH3)n]2+ Ions (n = 0−4)a

Figure S4). Obviously, only one parameter set is correct. In refs 8 and 9, the formula of the compound was not correct, and, not surprisingly, the density obtained from the lattice parameters did not match the experimental value we obtained for compound 4. The parameters of the monoclinic cell derived in ref 9 resemble our ones, but not all diffraction peaks could be reproduced with the former set. Photocatalytic Activity of [NH4Cu(OH)MoO4] (4). Pal and co-workers2a tested [NH4Cu(OH)MoO4] (supposed to be (NH4)2[Cu(MoO4)2]) as a photocatalyst in degradation of organic dyes, which are dangerous environmental pollutants due to their toxicity and because they are not biodegradable. Their test molecule was a carcinogenic direct azo dye, Congo Red. They used visible light and compound 4 with various morphologies and found that the dye decomposes. To confirm the photoactivity of 4, we applied UV light in deaerated aqueous solution (2 × 10−5 M dye, 0.03 wt % catalyst, 375 nm wavelengthsee Supporting Information Figure S5 for the spectrum of the UV lamp). The kinetic curves are shown in Supporting Information Figure S6. To get a quantitative measure of the reaction rate, we calculated the pseudo-firstorder rate coefficients by a linear fit to the −ln(A/A0) − t curves (see Table 7).

λmax 20a−d

Species

Cu(NH3)42+ Cu(NH3)3(H2O)2+ Cu(NH3)2(H2O)22+ Cu(NH3)(H2O)32+ Cu(H2O)42+ a

Calcd

Found

Compound (our expt.)

581 621 678 752 790

583 623 694 755 800

[Cu(NH3)4]MoO4 [Cu(NH3)3(H2O)]MoO4 [Cu(NH3)2(H2O)2]MoO4 [Cu(NH3)(H2O)3]MoO4 [Cu(H2O)4]SO4·H2O

Measured on the [Cu(H2O)4]SO4·H2O sample.

This means that in the 1 to 4 conversion, the first major stable intermediate (compound 5) is formed by the substitution of three ammonia ligands by H2O molecules. The ammonia to water exchange can be expected to take place stepwise, going through partially NH3/H2O ligand-exchanged intermediates. The NH3 to H2O ligand exchange in the solid state is rather surprising because it is the opposite to what happens in aqueous solutions of hydrated copper(II) complexes, where, due to binding more strongly to the copper(II) ion than H2O, ammonia substitutes the coordiG


Article

Inorganic Chemistry nated water molecules.21 The curiosity is that the presence of water vapor induces a solid-state reaction. (b) Considerations on the Ammonia−Ligand Water Exchange in [Cu(NH3)4MoO4] (1). To reveal the mechanism of the process, we concentrate first on the primary step. A straightforward assumption could be that first ammonia is released, and a coordinatively unsaturated copper(II) environment is formed, which then coordinates water from the gas phase to fill in the vacancy in its coordination sphere. If this mechanism operates, then the process could be promoted by removing the ammonia from the gas phase over the sample and supplying water vapor to complete the process. To test this possibility, we placed the dry compound 1 into a closed crucible attached to a flask filled with dilute (0.5 M) sulfuric acid, acting as a water vapor source and ammonia absorber. No signs of the expected transformation into compound 5 have been observed in either case within 48 h, even upon gentle heating. Allowing that the transformation may take place on a time scale of several months, the postulated mechanism cannot produce 5 from 1 in 2 days as observed in humid air, so ammonia release as the first step of the mechanism, followed by absorption of H2O from the air, can be excluded. That something else happens is supported by our additional experiments: We placed [Cu(NH3)4]MoO4 into a vacuum flask containing liquid water, without contacting the liquid, and removed the air, this way increasing the partial pressure of water vapor. Light blue color appeared in minutes already at room temperature. It is reasonable to assume that ammonia could not be lost from the solid compound in such a short time (in dry state compound 1 in vacuum is stable), suggesting that the first step in the mechanism is sorption of H2O. By monitoring the Raman (Figure 5) and the diffuse reflectance UV spectra of the sample during water vapor uptake (Suppl. Figure S7), signatures of the expected intermediates have been detected. In particular, upon inspection of the Raman spectra of the sample during the 1 to 5 conversion (Figure 5), a series of peaks in the region of the ν1 symmetric Mo−O stretching mode can be observed that are between those of the reactant 1 and the product 5. The intensity of the ν1 Raman band of the starting [Cu(NH3)4]MoO4 (1) (876 cm−1) decreases in time, and three other high intensity (A1, ν1) ones appear, none of which coincides with that of compound 5 (ν1 = 905 cm−1). The new bands might be assigned to the tetracoordinated cis/ trans-[Cu(NH3)2(H2O)2]MoO4 (6) and [Cu(NH3)3(H2O)]MoO4 (7) or pentacoordinated intermediates containing a surplus water molecule. (c) Theoretical Considerations on Ammonia−Water Ligand Exchange in [Cu(NH3)4]MoO4 (1). To explain these observations, we propose the following mechanism for the sequential substitution of NH3 with H2O starting from [Cu(NH3)4]MoO4 (1). (1) Water is chemisorbed by compound 1 by the formation of a five-coordinated intermediate ([Cu(NH3)4(H2O)MoO4]). This step requires the presence of water in the gas phase above compound 1 at as high partial pressure as possible. Since the initial geometry of 1 is square planar, H2O as the fifth ligand will most probably attach to the copper ion in an axial position, and the initial geometry of the complex formed will first be square pyramidal, SP-5. Numerous geometrical isomers (SP-5 and TB5) and conformers are possible, see below. The distortion of geometry is characterized by the τ index described by Addison et al.22,27

Figure 5. Raman spectra of complex 1 (a), complex 5 (b), the partially deammoniated intermediates (6−7) (c) and in the molybdate vibration bands region.

(2) The complex formed, [Cu(NH3)4(H2O)MoO4] (where the axial ligand is set in bold) is rather flexible, with many, energetically hardly distinguishable conformers (shown by density functional theory calculations). One can assume that the bond between copper and the ligand in the axial position is the weakest one (reflected in the longer axial than equatorial bond) so that the breakage of the axial bond will be the easiest. If the interconversion of the isomers/conformers is rapid, all H


Article

Inorganic Chemistry

Table 9. Ab Initio Cu−O and Cu−N Distances of Pentacoordinated Hydrated Complexes Formed by Water Addition to [Cu(NH3)4]2+ and [Cu(H2O)3(NH3)]2+ Cations Cu−O, Cu−N distance /Å Central geometry

Ligand arrangement

Bond length relationship

Lost ligand

Formed ion

SP-5

Cu(NH3)4−OH2,ax

Cu(NH3)4(H2O) Isomers Oa ≫ Ne1,3,Ne2,4 H2O Cu(NH3)42+

SP-5 TB-5

Cu(NH3)3(H2O)−NH3,ax Cu(NH3)3−OH2,axNH3,ax

Ne1 ≫N a,Oa, Ne1,Ne3

TB-5

Cu(NH3)2(OH2) NH3,axNH3,ax

SP-5 SP-5

Cu(OH2)4−NH3,ax Cu(OH2)3(NH3)−OH2,ax

TB-5

Cu(OH2)2(NH3) OH2,axOH2,ax Cu(OH2)3−NH3,ax−OH2,ax

TB-5

NH3

τSP‑5=0, τTB‑5=1

Axial

Equatorial

Oa = 2.212

Ne1,3 = 2.087 Ne2,4 = 2.088

0.15

Ne1 Ne1 Ne2 Ne2 Ne3 Ne3

= = = = = =

2.154 2.247* 2.110 2.082* 2.155 2.082*

0.52

Ne1 Ne1 Oe2 Oe2 Oe3 Oe3 Oe4 Oe4

= = = = = = = =

2.023 2.028* 2.018 2.001* 2.018 2.001* 1.982 2.007*

0.15

No stable conformer found Cu(NH3)3(OH2)2+ Na = 2.063 Na = 2.055* Oa = 2.061 Oa = 2.091*

No stable conformer found Cu(H2O)4(NH3) Isomers No stable conformer found Oa ≫ Ne1, Oe2,Oe3,Oe4 H2O Cu(OH2)3(NH3)2+ Oa = 2.146 Oa = 2.152*

No stable conformer found Oe1 ≫ Na,Oa,Oe2,Oe

H2O

five ligands can assume the axial position, making it a candidate for departure. If H2O leaves, it will maintain the equilibrium with the gas phase. If, however, NH3 goes away, the result will be a chemical reaction, a formal NH3 to H2O substitution in an SP-4 structure. What makes this mechanism feasible is the fact that the d9 copper(II) ion orbital symmetries permit the occurrence of Berry’s pseudorotation.23a,b This is a phenomenon in which small distortion of a pentacoordinated structure leads to the switch of a ligand from axial to equatorial and some others from equatorial to axial position, without breaking any coordinative bond. For example, a trigonal bipyramid with a vertical axis formally “rotates” to one in which the axis is horizontal because the axial bond lengths decrease and two equatorial bond lengths increase and simultaneously the angle between the former two switches from 180° to 120° and that between the originally equatorial bonds changes from 120° to 180°. During the conversion, the molecule passes a close to square pyramidal geometry. The energy change during the transformation is minimal (as confirmed by our density functional theory calculations), leading to a conformational equilibrium involving many slightly different SP5 and TB5 geometries. This way, the [Cu(NH3)4(H2O)MoO4] intermediate isomers might transform into each other. In one of the isomers, [Cu(NH3)3(H2O)(NH3)MoO4], the ammonia ligand is in the axial position, with an extended Cu−N bond length. Through this isomer, the molecule can “leak out” of the conformational equilibrium by losing the NH3, which leads to the permanent incorporation of a water molecule into the coordination sphere. The tetracoordinated [Cu(NH3)3(H2O)]MoO4 (7) formed in this step has been

Cu(OH2)3(NH3)2+

Na = 2.029 Oa = 2.007

Oe1 = 2.153 Oe2 = 2.001 Oe3 = 2.000

0.56

detected experimentally during decomposition of compound 1 (λmax = 623 nm, Table 7). This conclusion is also supported by the observed reduction of the intensity of the Cu−N stretching (445 cm−1) and the emergence of the v(Cu−O) stretching (422 cm−1) Raman bands.19a,b (3) Compound 7 can pick up another water molecule (probably in axial position again), and the pentacoordinated structure can eventually get into the [Cu(NH 3 ) 2 (H2O)2(NH3)]MoO4 structure via the pseudorotation-like procedure described above. This intermediate loses again the weakly bound ammonia with the formation of [Cu(NH3)2(H2O)2]MoO4 (cis and/or trans). This complex can, again, pick up a water molecule from the gas phase, forming eventually [Cu(NH3)(H2O)3(NH3)]MoO4. When this intermediate loses ammonia, [Cu(H2O)3(NH3)]MoO4 (compound 5), the final product is formed. The experiments show that the reaction stops after three of the four ammonia ligands of 1 are substituted. (We note here that compound 5 transforms further on a time scale of many days, not via ligand exchange, but via formation of compound 4.) [Cu(H2O)3(NH3)]MoO4 (5), in principle, could also pick up an H2O molecule, which, eventually, could replace the last NH3 ligand, which seems not to occur. It is not trivial why the last NH3 is not substituted. In order to find out the reasons for this, we performed electronic structure calculations on the SP-5 and TB-5 geometries of the two 4:1 intermediate complex cations [Cu(NH3)4(H2O)]2+ and [Cu(H2O)4(NH3)]2+ formed by water addition from compounds 1 or 5, respectively. I


Article

Inorganic Chemistry The purpose was to see the relative chances of the loss of the minority ligand, based on the bond lengths. The idea is that the length and the strength of a bond correlate: the longer a bond, the weaker it is. The lengths of the Cu−N and the Cu− O bonds are very close to each other in these complexes, so the relative binding energy of the NH3 and H2O ligands can be estimated by comparing the respective bond lengths. The identified conformers as well as the metal−ligand bond lengths are listed in Table 9. It is worth noting that the energies of the ions at numerous geometries have been evaluated during the search for stable conformers. The calculated values very rarely fell outside an about 5 kcal/mol range around the average values corresponding to each ion. As can be seen from Table 9, for both [Cu(NH3)4(H2O)]2+ and [Cu(H2O)4(NH3)]2+, one stable SP-5 (with τ = 0.15) and one TB-5 conformer (τ = 0.52 for the tetraammine and τ = 0.56 for the tetraaqua complex) were found. For both ions, in the SP-5 conformers the axial position is occupied by the H2O ligand, and in the TB-5 structures one axial ligand is an H2O and one is an NH3 ligand. Note that the TB-5 structures are quite far from the ideal C3 symmetry. First, we consider the chances of ammonia loss from [Cu(NH3)4(H2O)]2+. In the SP-5 conformer, as expected, the axial Cu−O bond is the longest, indicating that H2O loss can be expected. In the TB-5 conformer, the longest bond is one of the Cu−N bonds, showing that this bond has the largest propensity to break. The difference is significant: the Cu−N bond is at least 0.15 Å longer than the longest Cu−O bond. This also means that from this ion, an NH3 loss is feasible. On the other hand, in both stable conformers of [Cu(H2O)4(NH3)]2+, the longest metal−ligand bond is a Cu−O bond. From this one can conclude that the leaving ligand is H2O. The NH3 ligand is strongly bound to the metal ion and is not susceptible to removal. This is the reason why the last ammonia ligand is lost only at much higher temperature than the first three ones in the sequential ammonia loss of [Cu(NH3)4]MoO4.6 (d) Interligand Proton-Transfer Reaction of [Cu(H2O)3(NH3)]MoO4 (5) Yielding NH4Cu(OH)MoO4 (4). When compound 5 is left in open air for several days, [NH4Cu(OH)MoO4], compound 4 is formed. The IR spectrum of this product does not display any bands characteristic to coordinated ammonia. In its UV−vis spectrum the 800 nm band in Figure 6 corresponds to tetracoordinated CuO4 structures and differs from those of the chromophore groups with Cu−N bonds (Table 9). Deuteration experiments showed (see below) that the formation of compound 4 from compound 5 is irreversible. (e) Deuteration Experiments of NH4Cu(OH)MoO4 (4) and the 1/D2O and 1-d12/H2O Reactions. Deuteration experiments of compound 4 with D2O did not produce any [Cu(D2O)3(ND3)]MoO4 (5-d9). None of the possible deuterated isotopologs of compounds 5−7 were detected even when boiling NH4[Cu(OH)MoO4] in D2O. To interpret this result, we note that H/D exchange is efficient if the ammonia ligands depart, undergo isotope exchange, and return. On the other hand, if the exchangeable H or D is in a strongly hydrogen-bonded environment within the crystal, the ligand cannot leave and no isotope exchange takes place. The lack of any deuterated product shows that this latter is the case with compound 4. This also explains why formation of 4 is irreversible: crystals of 4 are cross-linked polymers, which are rather inert.

Figure 6. UV spectra of [Cu(NH3)4]MoO4 (1), [Cu(NH3)(H2O)3]MoO4 (5), [NH4Cu(OH)]MoO4 (4), and [Cu(H2O)4]SO4·H2O.

Formation of 4 from 5 is formally a self-protonation reaction: the coordinated ammonia ligand of 5 receives a proton from one of the coordinated water molecules and simultaneously releases the other two water molecules. The possible mechanisms are a quasi-intramolecular protonation first, followed by water loss, or vice versa (see Scheme 1), but, Scheme 1. Possible Steps of the (Self)Protonation Reaction during Formation of Compound 1

in fact, the proton may arrive from a different copper center, too. Isotope exchange experiments can help in revealing which mechanism operates. Accordingly, 1-d1228 was subjected to H2O vapor (as described above, compound 1 reacts with H2O vapor to give compound 4). If the protonation is quasi-intramolecular, then the last remaining ND3 is protonated by the H2O ligand coming from the surrounding air. Then exclusively NHD3[Cu(OH)MoO4] should be formed: [Cu(ND3)4 ]MoO4 + H 2O = ND3H[Cu(OH)MoO4 ] + 3ND3

because, once precipitated, NHD3Cu(OH)MoO4, just as NH4[Cu(OH)MoO4] seen above, cannot take part in H/D exchange with the solvent H2O. The IR spectrum of the reaction product, however, contained only NH4+ and OH bands without any sign of deuterium substitution. This shows that the [Cu(NH3)4−x(H2O)x]MoO4 intermediates are active in deuterium isotope exchange, which supports the mechanism of ligand substitution mentioned above: the pentacoordinated complexes are in dynamic equilibrium of multiple ligand exchange steps, with ligands coming and going. What is the J


Inorganic Chemistry

■

sink in the balance is the irreversible formation of the polymer of 4. In light of this observation, it looks as if the deuteration experiments could not answer which one of the two postulated mechanisms of the ammonia protonation operates. There is, however, another source of information that helps us to demonstrate that the reaction follows the pathway on the lefthand side of Scheme 1. Namely, keeping compound 5 at room temperature for a week, it partially decomposes. Sharp lines corresponding to coordinated water (1638 and 1618 cm−1), ammonium-ion (1409 cm−1, as well as 2022, 2806, 3012, 3235, and 3413 cm−1) and hydroxide-ion (3478 and 3552 cm−1) can be observed in the IR spectrum of the partially transformed 5 (Scheme 1) (Figure 7); the bands characteristic for the coordinated ammonia are missing.

Article

EXPERIMENTAL SECTION

In the synthesis experiments chemically pure sodium and ammonium molybdate, copper(II) sulfate pentahydrate, copper(II) oxide, molybdenum trioxide, and 25% aqueous ammonia solution (Deuton-X Ltd.) were used. Chemical grade of CaO, ammonium chloride, and pro. anal quality abs. ethanol, diethyl ether, glacial acetic acid, NH4OAc, HCl, NaOH, NaCl, disodium EDTA, NaHCO3, and 8-hydroxyquinoline (Deuton-X Ltd., Hungary) were also used. Determination of Molybdenum and Copper Content. A ca. 50 mg of Mo-containing sample was weighed and 30 mL of buffer solution (3 vol. of 50% ammonium acetate and 4 vol. of 50% acetic acid) was added, and then 30 mL of 0.5 m Na2EDTA solution was added to complete the dissolution of the sample. The solution was diluted to 80 mL and then heated until boiling; then 20 mL of 3% (wt) solution of 8-hydroxyquinoline dissolved in glacial acetic acid was added. The mixture was boiled for 3 min and filtered (G4 glass filtered); the yellow precipitate (MoO2(oxinate)2) was washed with hot water and dried at 135 °C until constant weight. The green solution was heated to 70 °C, 20 mL of oxine solution was added and then the pH was adjusted to be slightly alkaline with adding conc. ammonia solution. The greenish yellow precipitate (Cu(oxinate)2) was filtered off, washed with hot water, and dried at 100 °C. Determination of Ammonia Content. Roughly 0.5 g of sample was weighted and dissolved or suspended in ca. 10 mL of water and then 20 mL of 5% aq NaOH was dropped through a Teflon-lined dropping funnel and the solution was heated until starting the water distillation. The light blue precipitate containing solution turns into a black suspension. The ammonia evolved was absorbed in 0.1 M HCl solution and the excess of hydrochloric acid was titrated with 0.1 M NaOH. Vibrational Spectroscopy. FT-IR spectra of solid samples were recorded in the attenuated total reflection (ATR) mode on a Bruker Alpha FT-IR spectrometer at 2 cm−1 resolution and a Biorad Excalibur Series FTS 3000 infrared spectrometer, in KBr pellets between 4000 and 400 cm−1. The Raman measurements were performed using a Horiba JobinYvon LabRAM-type microspectrometer with an external 532 nm Nd:YAG laser source (∼40 mW) and an Olympus BX-40 optical microscope. The laser beam was focused by an objective of 10×. A D1 intensity filter decreased the laser power to 10% to avoid the thermal degradation. The confocal hole of 1000 μm and 1800 groove mm−1 grating monochromator were used in a confocal system and for light dispersion. The spectral range of 100−4000 cm−1 was detected as the relevant range with 3 cm−1 resolution. Each spectrum was collected at 240 s per point. XPS Measurements. X-ray photoelectron spectra were recorded on a Kratos XSAM 800 spectrometer operated at fixed analyzer transmission mode using Mg Kα1,2 (1253.6 eV) ) and Al Kα1,2 (1486.6 eV) excitation. The pressure of the analysis chamber was lower than 1 × 10−7 Pa. Wide scan spectra were recorded for all samples in the 100−1300 eV kinetic energy range with 0.5 eV steps and a 0.5 s dwell time. High-resolution spectra of the characteristic photoelectron lines of the main constituent elements, Cu 2p, O 1s, N 1s, Mo 3d, and the C 1s contaminations, were recorded by 0.1 eV steps and minimum 1 s dwell time. Spectra were referenced to the binding energy (BE) of the C 1s line of adventitious carbon, set at 284.6 eV. Quantitative analysis, based on peak area intensities (after removal of the Shirley type background), was performed by the Vision 2 and XPS MultiQuant programs using experimentally determined photoionization cross-section data of Evans and asymmetry parameters of Reilman.24a−c UV−Vis Measurements. Diffuse reflectance spectra in the UV− Vis region were detected at ambient by a Jasco V-670 UV−vis spectrophotometer equipped with an NV-470 type integrating sphere using the official BaSO4 standard as a reference. Powder X-ray Diffractometry. X-ray powder diffraction measurements were performed using a Philips PW-1050 Bragg− Brentano parafocusing goniometer. It was equipped with a Cu tube

Figure 7. IR spectra of the intermediate phase formed during partial decomposition of [Cu(H2O)3(NH3)]MoO4 into [NH4Cu(OH)MoO4].

After complete conversion to 5, the water lines disappear. This unambiguously shows that the formation of 4 takes place via the [NH4Cu(OH)(OH2)2MoO4] intermediate (see the left-hand side of Scheme 1. The first step is the selfprotonation, followed by departure of the water molecules from the coordination sphere of the Cu ion. Thermal Transformations of [Cu(NH3)4]MoO4 (1) and [Cu(NH3)(H2O)3]MoO4 (5)Formation of [Cu(NH3)MoO4] (8). Upon isothermal heating at 155−160 °C, both compounds 1 and 5 decompose, both yielding the same product. From 1 three ammonia and 5 three water ligands are eliminated, with the formation of stable [Cu(NH3)MoO4] (8). There was no reaction of the strongly bound ammonia ligand in solid 8 with H2O vapor and liquid water at room temperature. This means that the formation of compound 4 from 1 does not take place by removal of three ammonia ligands producing 8 as an intermediate, which is then hydrated with water vapor at room temperature. When, however, 1 is boiled in water, then, together with compound 4, an unassigned phase has been detected by powder XRD. We think that this phase can be the coordinated ammonia-containing intermediate, which is supported by the FT-IR data. K


Article

Inorganic Chemistry operated at 40 kV and 35 mA tube power, a secondary beam graphite monochromator, and a proportional counter. Scans were recorded in step mode. Evaluation of the diffraction patterns had been obtained by full profile fitting techniques. Thermal Studies. Thermal data were collected using a TA Instruments SDT Q600 thermal analyzer coupled to a Hiden Analytical HPR-20/QIC mass spectrometer. The decomposition was followed from room temperature to 780 °C at 5 °C min−1 heating rate in argon or air as carrier gas (flow rate = 50 cm3 min−1). Sample holder/reference: alumina crucible/empty alumina crucible. Sample mass ∼ 7 mg. Selected ions between m/z = 1−120 were monitored in multiple ion detection mode (MID). Photocatalytic Studies. Photocatalytic activity was investigated by putting 30.0 mg of the sample into a 100 cm3 aqueous solution of Congo red dye (2 × 10−5 M) in a double-walled borosilicate glass reactor. The solution absorbance was measured with and without the photocatalyst (compound 4). The reactor was covered to prevent evaporation, stirred continuously, and cooled with flowing water during the experiment. After stirring for 1 h in the dark for the adsorption equilibrium to occur, six Osram BLACKLIGHT BLUE 18W/73 G13 UV lamps with a distance of 5 cm from the reactor were used for UV light irradiation. At every hour for 4 h, 3 cm3 samples were taken from the solution with an automatic pipet, centrifuged, and the decomposition of the Congo red dye was followed by measuring the absorbance of its most intensive peak at 497 nm with a Jasco V-550 UV−vis spectrophotometer. Electronic Structure Calculations. Structures of the pentacoordinated complex ions have been calculated using density functional theory. The geometries optimized at the M05-2X/LANL2DZ and the CAM-B3LYP/LANL2DZ(Cu)+6-31G**(H, C, N, O) levels agree very well. The differences between the bond lengths of Cu−N and Cu−O bonds in axial and equatorial positions obtained at the two levels of theory are essentially the same. Table 9 contains the bond lengths calculated at the M05-2X/LANL2DZ level. SEM Measurement. Scanning electron miscroscopy measurements were performed by a Zeiss EVO40 microscope at 20 kV. Synthetic Procedures. Compound 1. (a) 5.0 g copper(II) sulfate pentahydrate (20 mmol) was dissolved in a minimal amount of distilled water and then 3.94 g (20 mmol) of ammonium molybdate ((NH4)2MoO4) dissolved in water was added. The formed turquoisecolored fine basic copper molybdate precipitate was filtered by suction and then washed thoroughly with distilled water until sulfate free. The wet precipitate was dissolved in conc. ammonium hydroxide, the opaque solution was filtered, then absolute ethanol was added to precipitate the purplish colored fine crystalline mass precipitate. The precipitate was filtered off and washed with ethanol.5 The compound was dried in a desiccator containing calcium oxide and solid NH4Cl to ensure ammonia atmosphere during the entire drying process. Found (in wt %): Cu-21.88; Mo-33.73; NH3-21.90. Calculated for [Cu(NH3)4]MoO4: Cu-21.78; Mo-32.93; NH3-23.32. Experiments performed to precipitate the complex with MeOH failed. (b) Following the method given in route a, the solution formed after dissolution of the basic copper molybdate in conc. ammonium hydroxide, instead of ethanolic precipitation, was filtered and cooled in a fridge to ∼0 °C and the formed small blue crystals were collected by filtration. The crystals were dried in a desiccator (CaO/NH4Cl). Found (in wt %): Cu-21.97; Mo-33.43; NH3-22.17. Calculated for [Cu(NH3)4]MoO4: Cu-21.78; Mo-32.93; NH3-23.32. (c) 7.9 g copper(II) oxide (0.1 mol) and 14.4 g of molybdenum trioxide were mixed and ground in an agate mortar, and then the finely ground mixture was heated at 630 °C for 2 h in an oven in air. The formed copper(II) molybdate (CuMoO4) was dissolved in 350 mL of conc ammonium hydroxide; the opaque solution was filtered and left to crystallize in a refrigerator for a day, when dark blue crystals of 1 were formed. The crystalline mass was dried in open air. Found (in wt %): Cu-22.06; Mo-33.17; NH3-21.76. Calculated for [Cu(NH3)4]MoO4: Cu-21.78; Mo-32.93; NH3-23.32. (d) Following the synthetic method c, the isolation of the product was done by mixing with ethanol instead of cooling in a refrigerator. The purplish, fine crystalline mass of 1 could be isolated and dried in

a desiccator over CaO in the presence of NH4Cl. Found (in wt %): Cu-21.99; Mo-32.54; NH3-21.70. Calculated for [Cu(NH3)4]MoO4: Cu-21.78; Mo-32.93; NH3-23.32. Compound 5. Compound 5 was synthesized by putting a sample of 1 into a vacuum desiccator coupled to a water-filled flask. The desiccator was evacuated, and the water vapor filled the desiccator. In a few hours compound 1 was turned into a light blue powder. Found (in wt %): Cu-21.92; Mo-33.14; NH3-5.12. Calculated for [Cu(NH3)(H2O)3]MoO4: Cu-21.56; Mo-32.60; NH3-5.77. Compound 8. Compound 8 was prepared by isotherm heating of compound 1 at 155−160 °C for 2 h. The light greenish yellow powder was analyzed for its Cu, Mo, and ammonia content. Found (in wt %): Cu-26.93; Mo-39.99; NH3-6.87. Calculated for [Cu(NH3)MoO4]: Cu-26.40; Mo-39.91; NH3-7.07. Compound 3. Ammonium paramolybdate (NH4)6Mo7O24·4H2O (8.55 g, 0.05 mol) and copper(II) nitrate trihydrate (12.1 g 0.05 mmol) were dissolved in a mixture of 50 mL of water and 50 mL of cc NH4OH. The dark blue solution formed was boiled until the solution became colorless and the ammonia smell deceased (

Gas Phase Ligand Exchange

Recommend Documents