Synthesis, Structure, and Actual Applications of Double Metal Cyanide Catalysts

Abstract

Double metal cyanide (DMC) complexes represent a unique family of materials with an open framework structure. The main current application of these complexes in chemical industry is their use as catalysts (DMCCs) of the ring-opening polymerization of propylene oxide (PO), yielding branched polyols, highly demanded in production of polyurethanes and surfactants. The actual problem of chemical fixing carbon dioxide from the atmosphere gave new impetus to the development of DMCCs, which turned out to be effective in oxirane/CO₂ copolymerization. In recent years, new types and formulations of DMCCs were created, so that greater understanding of the reaction mechanisms was achieved and new fields of catalytic applications were found. In the present review, we summarized background and actual information about the synthesis, structure, and mechanisms of the action of DMCCs, as well as their application in the development of new materials and fine chemicals.

1. Introduction

Commodity, energy, and environmental concerns drive scientists to develop contemporary approaches to new materials, based on the methods of green chemistry and utilization of wastes and greenhouse gases; improving the efficiency of oil processing is also on the agenda [1,2,3]. Among new-old catalytic systems, widely used earlier for conversion of petroleum-based monomers to in-demand polymer materials and that have received new life in chemical fixation of CO₂ [4], double metal cyanide (DMC) complexes hold a special place among heterogeneous catalysts [5].

The increased interest in these catalysts arises from ability of DMC complexes to initiate ring-opening polymerization (ROP) of oxiranes (Scheme 1a), mainly propylene oxide (PO), with a formation of polyethers that have multitude of applications ranging from soft components in polyurethane foaming to pharmaceutical formulations [6,7]. The ability of DMC catalysts (DMCCs) to introduce CO₂ to growing polymer chains with the formation of poly(ether-carbonate)s (PECs) (Scheme 1b) [8,9] maintains interest in these catalytic systems, forcing scientists to continue studies in this field. In recent reviews, the matters of the synthesis, structure, and reaction mechanisms of DMCCs [5,10,11,12,13] and their applications in polymer chemistry [4,5,6,7,8,9,14,15] were discussed. However, the topic of DMC complexes is not exhausted by these aspects, and in the present work we tried to describe and discuss all facets of the preparation, catalytic action, and use of DMCCs, focusing on specific applications, new processes, and materials, based on recent publications in this field.

2. Preparation and Structure of DMC Catalysts

DMC catalysts represent heterometallic complexes comprising metal–cyanide and metal–ligand fragments of the formula M_a[M′(CN)_b]_c·xMX_d·yL·zH₂O, where M = Zn, Fe(II), Ni(II), etc.; M′ = Co(III), Fe(II), Fe(III), etc.; MX_d—a salt of metal M, used in the synthesis of DMC complex; and L—organic ligand(s). Stoichiometric factors x, y, and z are determined by the method of preparation as well as type and ratio of the reagents [10,16]. Among DMC complexes, Zn(II)-Co(III) (Zn/Co) systems are the most accessible and widely used. Since 1966, when the use of Zn(II)-Fe(III) (Zn/Fe) and Zn/Fe as a catalysts for polymerization of propylene oxide (PO) was proposed [17], the methods of the preparation of DMCCs were improved, and the chemical nature, composition, and structure of DMC complexes were studied. In scientific periodicals, DMC complexes were first described in 1973 [18], and the first study on optimization of Zn(II)-Fe(III) and Zn(II)-Co(III) catalysts was conducted in the mid-1980s [19]. Comprehensive, reproducible, and reliable results of the studies in the field of DMCCs began to appear in early 2000s, which was due to the progress in physico-chemical methods of the analysis of heterogeneous catalysts and polymers, as well as the development of the theory of catalytic processes with the use of quantum-chemical modeling [20,21,22,23,24,25,26,27,28,29]. In this Section, we present the basic and later results of the studies of preparation and structure of DMC complexes, both widely used Zn/Co systems and prospective bimetallic complexes with different ligands.

2.1. Preparation of DMC Complexes

2.1.1. Synthesis of Zn(II)-Co(III) DMCCs

The main method of the preparation of Zn/Co DMC complexes is based on exchange reaction between cyano complexes of Co(III) and Zn(II) salts in the presence of water and (optionally) donor organic ligands (complexing agents, CAs, Scheme 2). Significant excess of Zn(II) salts (usually ZnCl₂) is critical in the preparation of DMCCs [30], and ^tBuOH is widely used as a CA.

A typical synthetic procedure of the preparation of highly active nano-sized Zn/Co DMCCs [20] comprises addition of the aqueous solution of K₃[Co(CN)₆] to the aqueous ZnCl₂ solution with subsequent addition of the solution of ^tBuOH and nonionic detergent in water, followed by separation of the reaction product by centrifugation, washing by boiling ^tBuOH, and drying. Another order of the mixing of reagent solutions can also be used, e.g., addition of the ZnCl₂ solution to the solution of ^tBuOH and surfactant with subsequent treatment with K₃[Co(CN)₆] [23]. An alternative method of the synthesis of Zn/Co DMCCs uses simultaneous addition of ZnCl₂ and K₃[Co(CN)₆] solutions to the solution of CA; this ‘parallel-flow dropping’ method [31] in some cases allows to obtain more efficient catalysts of copolymerization of PO with CO₂ (see Section 4.2). Active catalysts, prepared using ZnCl₂ and ^tBuOH, have a common formula, Zn₃[Co(CN)₆]₂·x(ZnCl₂)·y(H₂O)·z(^tBuOH). These catalysts are raw material intensive; for example, to prepare about 10.0 g of a DMCC, 20.0 g of ZnCl₂, 8.0 g of K₃[Co(CN)]₆, 205 g of water, and 410 g of ^tBuOH are needed [32]. To control the reaction, in situ Raman spectroscopy was recently employed. It was shown that at the final stage of the formation of Zn/Co DMCC by the reaction of Zn₃[Co(CN)₆]₂ intermediate with ZnCl₂ and ^tBuOH, the reaction time could be adjusted from hours to minutes through the control of reactant concentrations [33]; however, the further studies on optimization of the Zn/Co DMCC yield are waiting for their researcher. Note that in all early articles [20,21,22,23,24,25,26,27,28], more recent studies [27,34,35], and patents, e.g., [36], the isolated yields of the catalysts were not pointed, making it difficult to compare the efficiency of different methods.

The sequence of mixing of the reagent solutions as well as type and quantity of CA(s) used in preparation of Zn/Co DMCCs directly affect the surface properties of materials. For example, the catalyst prepared by co-precipitation of ZnCl₂ and K₃[Co(CN)₆] solutions in the absence of CAs had a specific surface area (SSA) of 895 m²∙g⁻¹, pore volume (PV) = 0.592 mL∙g⁻¹ and an average pore diameter d = 1.3 nm; whereas, depending on the mixing order, reaction temperature, and time, the samples of Zn/Co DMCCs containing ^tBuOH and poly(ethylene glycol) (M_n = 4000, PEG-4000) had SSA = 48.3–486 m²∙g⁻¹, PV = 0.251–0.518 mL∙g⁻¹, and d = 2.1–10.3 nm [37].

The use of ZnCl₂ as a source of Zn²⁺ ions provides higher activity of Zn/Co DMCCs in comparison with the catalysts prepared using other Zn(II) halogenides [27,34,35], Zn(OAc)₂ [38,39], ZnSO₄ [38,40,41], or Zn(CF₃SO₃)₂ [40,41]. DMCCs synthesized at higher ZnCl₂/K₃[Co(CN)]₆ ratios demonstrate higher catalytic activity [42]. As was shown by Qi et al. [43], highly active Zn/Co DMCCs can be prepared by the reaction of H₃[Co(CN)₆] with Zn(OH)₂, basic ZnCO₃, or ZnO in ^tBuOH/H₂O, followed by suspending in aq. HCl and washing by ^tBuOH (the yields were not given), but by its nature this catalyst can hardly have been different from ZnCl₂-based Zn/Co DMCCs. However, simple mixing of H₃[Co(CN)₆] and ZnCl₂ in MeOH resulted in precipitation of [ZnCl]₂[HCo(CN)₆]∙MeOH with near quantitative yield [44].

In addition to ^tBuOH, different polar organic compounds (e.g., crown ethers [45,46], dicarboxylic acids [47]) and polyethers [20,48,49,50,51,52] were used as co-CAs in Zn/Co DMCC formulations. Organic detergents and dispersants also find use in the preparation of Zn/Co DMCCs; for this purpose, polyacrylamides [53], polyethoxylated sorbitan oleate (Tween^® 80) [54], and other reagents [10] were applied.

Besides ^tBuOH, many other organic compounds (alcohols, aldehydes, ketones, ethers, esters, carboxylic acids, amides, ureas, nitriles) were tested as CAs [7,10,55]. In an early study [18], the use of Ca₃[Co(CN)₆]₂ was proposed in combination with 1,2-dimethoxyethane; the yield of Zn/Co DMC complex was specified in this work, amounting to 1.05 g of the catalyst per 1 g of the starting Ca₃[Co(CN)₆]₂.

Replacement of ^tBuOH by eco-friendly lactate esters in the synthesis of Zn/Co DMCCs was proposed by Kim et al. in 2011 [32] (the yields were not given), the catalysts obtained from HOCHMeCOOⁱPr and polyethers as CAs were not inferior on productivity to ^tBuOH-based DMCCs (see Section 4.1.1). A comparative study of the use of ^tBuOH and dicarbonyl compounds of different structure (Scheme 3) in preparation of Zn/Co DMCCs, conducted in 2019 by Kim et al. [52], revealed different types of coordination of CAs (see Section 2.2.1) affecting the activity of the catalysts. The yields of these catalysts were ~1.0 g based on 0.7 g of K₃[Co(CN)₆]. The range of dicarbonyl compounds, used as CAs in preparation of Zn/Co DMCCs, was extended in later studies [56,57]; α-, β-, γ-, and δ-dicarbonyl compounds (dicarboxylic acids, dicarboxylic esters, ketoesters and diketones) were investigated, and conclusions about the preference of keto-coordination of CAs for higher catalytic performance in PO polymerization [57] and PO/CO₂ copolymerization [56] received further experimental proof. Simple and “greener” alternatives to dicarbonyl compounds were cyclic carbonates, proposed by Kim et al. in 2024 to replace ^tBuOH in Zn/Co DMCC formulations; the catalysts were prepared in the presence of P-123 as a co-CA, and the yields were not given [58]. In 2024, active Zn/Co DMCCs were also prepared with the use of dialkyl phosphonates, trialkyl phosphites, and trialkyl phosphates as CAs [59]; the yields were not provided.

In 2019, preparation of crystallizable Zn/Co DMCC with the use of Zn(OAc)₂ instead of ZnCl₂ was reported [60]. The obtained complex Zn₂[Co(CN)₆](OAc)∙4H₂O comprised positively charged {Zn₂Co(CN)₆}⁺ DMC layers linked through OAc groups (XRD data). In some catalytic processes (see Section 4), this complex was not inferior to conventional Zn/Co DMCCs, prepared with the use of ^tBuOH.

The idea of the use of H₃[Co(CN)₆] in the synthesis of Zn/Co DMCCs [43] was painted with new colors when Lee et al. set themselves the task of the preparation of well-defined DMCCs. In 2021, they prepared adducts of H₃[Co(CN)₆] with EtOH and MeOH and synthesized a Zn/Co DMC complex by the reaction of H₃[Co(CN)₆]∙2.4EtOH (weight form) with Zn(II) 2-ethylhexanoate [61]. In 2024, they synthesized a series of complexes, described by the formula [XZn⁺][Zn²⁺][Co(CN)₆³⁻] where X = Cl, ^tBuO, and/or AcO in different ratios [62]. The methods of preparation of these well-defined Zn/Co DMCCs are presented in Scheme 4.

The use of nitriles [63,64], amides [65], or nitromethane [65] as CAs for Zn/Co DMCCs was studied recently by the scientific group of Kim. The same scientific group also investigated the use of silyl esters as an alternative to ^tBuOH [66]. Verma et al. proposed the use of EDTA as a CA (however, complexing properties of this chelating ligand seemed a bit redundant) [67]. The yields of the catalysts were not given in these works.

Low-MW aliphatic amines showed low efficiency as CAs, but higher aliphatic amines demonstrated promising properties when used in preparation of Zn/Co DMCCs [68]. High activity of the catalyst can be attributed to excellent morphology of the catalyst particles, apparently caused by the action of higher amines as surfactants. In [69], cetyltrimethylammonium bromide was used as a CA in the synthesis of Zn/Co DMCC (no yield data) of the selective formation of cyclic carbonates (see Section 4.6).

To reduce solvent and reagent consumption, mechanochemical approaches to Zn/Co DMCCs were examined. In 2011, Zhang et al. proposed a liquid-assisted grinding method for the preparation of Zn/Co DMC complex from ZnCl₂ and K₃[Co(CN)₆] with the use of ^tBuOH/decafluoropentane mixture as a grinding liquid (no yield data) [70]. Ball milling was also used for pre-treatment of Zn/Co DMC before its use in preparation of three-component catalyst for copolymerization of PO, CO₂, and trimellitic anhydride [71] (see Section 4.3).

One of the modern directions of the design of Zn/Co DMCCs is an increase of their surface area via preparation of supported catalysts. In particular, MCM-41-supported catalysts were prepared by simultaneous addition of ZnCl₂ and K₃[Co(CN)₆] solutions to MCM-41 suspensions containing different CAs [72]. Dispersing of Zn/Co DMCC on silica by dry or liquid-assisted grinding was also studied, but a substantial part of the catalyst remained unsupported [73]. More efficient binding of Zn/Co DMCC particles with silica support was achieved when DMC complex was co-precipitated with silica at moderately acidic conditions (sol–gel method) via addition of aq. K₃[Co(CN)₆] to an aqueous solution containing ZnCl₂, Si(OEt)₄ and HCl, followed by heating until a gel formation, gel separation, drying, and grinding [74]. To increase Lewis acidity and catalytic activity of the DMCCs, a similar method was proposed for the synthesis of Zn/Co DMCCs@TiO₂ with the use of Ti(OEt)₄ instead of Si(OEt)₄ [75]. To increase the catalytic activity of the Zn(OAc)₂-based layered DMC complex [60], it was supported in situ on Al-Beta-25 zeolite [76].

To sum up, there are two common synthetic approaches to Zn/Co DMCCs. The first approach is based on ion exchange between K₃[Co(CN)₆] and ZnCl₂ in the presence of CA and dispersant. As was shown recently by Chen et al. [33], Zn₃[Co(CN)₆]₂ intermediate is formed rapidly, but further formation of the catalyst takes place over time. Non-stoichiometry of the formation of DMCCs and variability of its composition do not allow a correct comparison of the different methods in terms of the catalyst’s yield. In most cases, the yield of Zn/Co DMCC (in wt%) is roughly equal to the mass of starting K₃[Co(CN)₆], but the consumption of ZnCl₂ (to provide surface saturation of the catalyst by Zn–Cl species) and ^tBuOH (to provide its introduction to the catalyst’s framework and elimination of the K⁺ ions) are disproportionately large. The second, more advanced approach to Zn/Co DMCCs is based on the use of K-free cyano complexes of Co [44,62]. This approach provides the formation of Zn/Co DMC complexes with a given composition and high isolated yields, with wide spectrum of catalytic applications. The comparison between the effectiveness of different methods of the preparation of Zn/Co DMCCs is significantly complicated by the absence of the data on the catalyst’s yields in the vast majority of works employing the co-precipitation method.

2.1.2. Synthesis of Zn(II)-Fe(III) DMCCs

Zn/Fe DMC complexes were synthesized and studied in polymerization of PO along with Zn/Co systems back in 1966 [17]. The methods of the preparation of Zn/Fe DMC are similar to those applied in the synthesis of Zn/Co DMCCs, and ^tBuOH was widely used as a CA in the preparation of these complexes. In the very recent work of Laguta and van Koningsbruggen [77], the formation of conventional Zn/Fe DMCCs with the use of ^tBuOH was studied using the approaches of colloid chemistry. They showed that the catalyst particle at the stage of its formation represents micelle-like species comprising Zn/Fe(CN)_x aggregate, decorated by Zn²⁺ ions, and Cl⁻ counterions form the Stern and diffuse layers. Based on the comparison of the Gutmann donor numbers (38 and 33 for ^tBuOH and H₂O, respectively) [77], the role of ^tBuOH in the synthesis was to act as a sterically bulky ligand to Zn²⁺, which shields zinc from the complex Fe(CN)_x ion.

Pd-doped systems with various amounts of Pluronic P123 co-CA are described in [78]. In the recent study of Kumar et al. [79], Zn/Fe DMCCs with Zn/Fe ratios of 1.49–2.93 were prepared with the use of Na₂EDTA as a CA, and PEG-200 was added as a co-CA. These catalysts contained minimal amounts of Cl⁻ but retained high catalytic activity in polymerization of oxiranes. As is the case of Zn/Co DMCCs, in most of the publications, the isolated yields of the catalysts were not given.

Relatively recent studies have aimed at finding more material-efficient approaches to these systems. Mechanochemical synthesis of Zn/Fe DMC complexes was performed for the first time by Zhang et al. in 2011 [70], and a simple liquid-assisted grinding method was employed. Solvent-free grinding was used for the preparation of Zn/Fe DMCCs from K₃[Fe(CN)₆] and ZnX₂ (X = Cl, OAc) [39]. Ball milling of the mixtures of K₃[Fe(CN)₆], ZnCl₂, ^tBuOH, and different co-CAs was used in preparation of Zn/Fe DMCCs employed in PO/CO₂ addition and copolymerization [80,81]. The mechanochemical approach to Zn/Fe DMCCs was modified by the use of imidazolium-based ionic liquids, which increased the thermal stability of the catalyst [82]. A highly active catalyst was also obtained by ball milling in the presence of quaternary ammonium salts [83]. The apparent advantages of mechanochemical approaches are high (near quantitative) yields of the catalysts and their known composition. However, when compared with conventional co-precipitation technique, mechanochemical methods provide a lower degree of phase homogeneity.

2.1.3. Synthesis of Other DMC Complexes

Synthetic approaches to other DMC complexes, different from Zn/Co and Zn/Fe(III) systems, are applying similar methods based on ligand exchange. To date, dozens of compounds have been synthesized and studied [84], and the synthesis and structure of similar systems, so-called Prussian blue analogs, have been discussed in recent reviews [11,12,13]. The main field of application of these complexes are the energy storage materials, which is not the subject of our study. In this Section, we will only address some fundamental works in this field and discuss recent publications concerning DMC complexes of interest for the science of catalysis.

A series of Co/M DMCCs were prepared by reacting of aq. CoCl₂ with K₂[Ni(CN)₄], K₄[Fe(CN)₆], K₃[Fe(CN)₆] or K₃[Co(CN)₆] in the presence of ^tBuOH and poly(tetramethylene ether) glycol (PTMEG) as the co-CA [85]. Porous DMC complexes Ni₃[Co(CN)₆]₂, Co₃[Co(CN)₆]₂, Fe₃[Co(CN)₆]₂, Ni₃[Fe(CN)₆]₂, Co₃[Fe(CN)₆]₂, and Fe₄[Fe(CN)₆]₂ were prepared by standard co-precipitation method using metal chlorides, hexacyanometallates of Co(III) or Fe (III), and ^tBuOH [86]. Hexacyanoferrate(II) derivatives of Zn, Fe(III), Co(II), and Ni(II) were similarly prepared [87]. Co-precipitation of the solutions of K₃[Co(CN)₆] and Cu(NO₃)₂ was used in the synthesis of Cu/Co DMC complex of the formula Cu₃[Co(CN)₆]₂∙11H₂O [88].

The synthesis of Zn/Ni DMCCs was conducted by the addition of aq. KCN to the solution of NiCl₂, followed by treatment with aq. Zn(OAc)₂ [89]. In [90], ZnCl₂ was used as a source of Zn for the preparation of this type catalyst. Co/Ni DMCCS were prepared by co-precipitation from K₂[Ni(CN)₄] and CoCl₂ in the presence of ^tBuOH and poly(THF) [91].

The complex Co(H₂O)₂[Pd(CN)₄]·4H₂O, with a clearly defined structure, was obtained by layering a solution of CoCl₂ in EtOH onto an aqueous solution of K₂[Pd(CN₄)], and other Co[M(CN)₄] complexes (M = Ni, Pt) were synthesized through combination of Co(II) salts with K₂[M(CN)₄]; in contrast with Zn/Co DMCCs, preparation of these DMC complexes required complete elimination of Cl⁻, which was a poison for active catalyst [92]. Mechanochemical methods were used in the synthesis of Zn/Cr DMC complexes (starting from K₃[Cr(CN)₆], ZnCl₂ and ^tBuOH in 1:10:0.1 molar ratio) [93].

2.2. Structure of DMC Complexes

The general structure of DMC complexes includes two metal centers coordinated at C and N atoms of CN⁻ ligands [7,52,94]. Detailed analysis of the structure of DMC complexes is complicated by their low solubility and by strong dependence of their composition and morphology on the method of synthesis. Non-stoichiometric character of active DMCCs [38] adds an additional level of complexity to the determination of the structure of these catalysts and makes it hard to understand the mechanism of the action of DMCCs.

In most works mentioned previously and devoted to the synthesis and use of DMC complexes, their structure and composition is determined by a standard set of methods, including:

• Elemental analysis to determine metal and Cl content;
• Scanning electron microscopy (SEM), including the use of energy-dispersive X-ray spectroscopy (EDX) to determine the size and morphology of the catalyst particles;
• X-ray diffraction (XRD) methods to determine the type of the crystal lattice and degree of crystallinity;
• FT-IR spectroscopy, which gives information about the chemical bonding (M–C≡N^…M’ fragments, CAs);
• Differential scanning calorimetry and thermogravimetry analysis (DCS/TGA), which reveals the quantity of CAs and the strength of their binding;
• New spectral methods such as extended X-ray fine structure/X-ray near edge structure (EXAFS/XANES) determination that allow determination of the ligand environment of the metal atoms;
• X-ray photoelectron spectroscopy (XPS), for the same purpose.

Also, as was mentioned above, at the stage of DMCC formation, Raman spectroscopy [33] and dynamic light scattering (DLS) [77] are useful and informative.

2.2.1. Structure of Zn/Co DMC Complexes

Well-defined structures of crystalline compounds Zn₃[Co(CN)₆]₂ and Zn₃[Co(CN)₆]₂∙12H₂O [95] have little in common with real structure of Zn/Co DMCCs. However, diversity of non-random vacancy arrangements, found in DMC complexes relatively recently [96], helps to explain high catalytic activity of DMCCs by the ease of mass transfer and defragmentation of catalyst particles observed during polymerization of oxiranes. There is a view that the structure of Zn/Co DMCCs can be considered as a combination of Zn₃[Co(CN)₆]₂ as the backbone and coordinatively unsaturated Zn²⁺ ions with Cl⁻, OH⁻ and ROH ligands on the surface, but this model is too simplistic [6]. Comparative study of Zn/Co DMCCs, prepared without and with CAs, revealed the presence of cubic crystalline phase in the first case, and formation of monoclinic phase in the latter case, also noting that some samples of the catalyst were amorphous [37]. In a recent study by Zhang et al. [97], the formation of a rhombohedral phase was detected in a Zn/Co DMC complex, prepared using ^tBuOH as a CA. The visible difference between a “pure” Zn/Co DMC complex and Zn/Co DMCC, prepared using MeCN as a CA, is presented in Figure 1 [64]. The DFT calculations of the face-centered cubic structure of “pure” Zn/Co DMC were based on a cubic Fm-3m symmetry model (Figure 1a). In contrast, for Zn/Co DMC-MeCN, a monoclinic P11m symmetry model was used (Figure 1c). These selections were made based on the SEM images showing the crystallite morphologies and the XRD patterns (Figure 1). The level of theory used in DFT optimization is not provided in [64].

The idealized structure of Zn/Co DMC complex, containing Cl⁻ ligands (Figure 2), corresponds to the chemical formula Zn₂[Co(CN)₆]Cl. In 2007, XRD studies of Zn(II)–Co(III) DMC complexes of different composition showed that pure Zn₃[Co(CN)₆]₂ had a crystallinity of 98.8% with a typical cubic lattice structure, while DMCCs prepared with excess ZnCl₂ and/or complexing agents had low crystallinity. It is noteworthy that the ZnCl₂ crystalline phase was not found in DMCCs prepared with an excess of ZnCl₂ (wide-angle X-ray scattering (WAXD) data), although it seemed that free ZnCl₂ existed in the catalyst (as evidenced by elemental analysis) [98]. In this work, experimental arguments in favor of Zn₂[Co(CN)₆]Cl as a possible base unit of Zn/Co DMCCs were found: the maximum of activity was observed at a 1:1 molar ratio of ZnCl₂ and Zn₃[Co(CN)₆] in DMC complex, which corresponds to Zn₂[Co(CN)₆]Cl stoichiometry. Also, it should be noted that when DMC complex Zn₂[Co(CN)₆]OH∙4H₂O was synthesized and studied back in 1987, it had a hexagonal crystal lattice structure [19], but corresponding atomic coordinates, used in further theoretical studies [29,99], were not made public.

FT-IR spectral analysis gives some insight into the nature of chemical bonds in Zn/Co DMCCs [31]. The FT-IR spectra exhibit characteristic absorption peaks of –CN, Co–CN and –OH fragments. Stretching vibration of CN⁻ in Zn₃[Co(CN)₆]₂·nH₂O appeared at 2178 cm⁻¹, which is shifted to higher wavenumber in comparison with ν(CN) of K₃[Co(CN)₆] (2128 cm⁻¹). It proves that CN⁻ acts not only as a σ-donor to Co³⁺ but also as an electron donor in the CN–Zn²⁺ fragment [37]. The shift of ν(Co–CN) band from 416 cm⁻¹ for K₃[Co(CN)₆] to 454 cm⁻¹ for Zn₃[Co(CN)₆]₂·nH₂O also proves the formation of a Co³⁺–CN–Zn²⁺ structure (Figure 3). In the CN⁻ adsorption region (Figure 3b), additional absorption overlapping peaks (~2190 and 2152 cm⁻¹) were detected in the spectra of the Zn/Co DMCCs; these peaks were attributed to amorphous phase and monoclinic structure, respectively. The absorption peaks of Co–CN at 445 and 473 cm⁻¹ were assigned to amorphous phase and monoclinic structure, respectively (Figure 3c) [31].

In 2015, a comprehensive study of the structure of Zn/Co DMCCs with the use of EXAFS/XANES determination, powder XRD, EDX, and XPS was conducted by Lawniczak-Jablonska et al. [100]. According to the results of this study, Zn/Co DMCCs, prepared with the use of ^tBuOH and excess ZnCl₂, do not contain ZnCl₂. EXAFS analysis showed that the coordination around Zn²⁺ was changed from octahedral in reference material Zn₃[Co(CN)₆]₂∙xH₂O to tetrahedral in catalysts, and Cl atoms were detected near some Zn atoms but no significant amount of O atoms was detected. The Zn atoms inside the clusters had an atomic order similar to anhydrous Zn₃[Co(CN)₆]₂ structure (tetrahedral coordination); the remaining Zn²⁺ ions were defined as bound partially with CN⁻ and with one or two Cl⁻, and located at the surface of the clusters without the formation of ZnCl₂ phase. The role of CAs is the formation of an amorphous phase, thereby increasing catalytic activity of the complexes [100]. Zn/Co DMCCs, prepared using 1,2-dimethoxyethane (DME), were also studied using EXAFS analysis, and Zn^…O coordination was not clearly detected [101]. Although EXAFS studies of Zn/Co DMCCs prepared using ^tBuOH or DME as CAs have not shown substantial contribution of direct Zn^…O coordination to the formation of Zn²⁺ ligand environment [102], DSC/TGA have revealed a binding between catalyst matrix and these CAs; it is noteworthy that DMCCs, prepared with the use of ^tBuOH, begin to decompose at 120 °C with a formation of isobutylene, which indicates the catalytic character of this process [102,103]. The formation of H₂O during thermal decomposition of Zn/Co DMCCs may change the nature of the catalyst via the formation of Zn–OH species, this process cannot be ignored in view of the conditions of reactions involving Zn/Co DMCCs. The mechanism of the catalytic action of DMCCs (see Section 3.1) is closely related to the structure of the complexes, and structural experimental data sometimes contradict hypothetical structures used as the basis for quantum-chemical modeling.

Comparative study of Zn/Co DMCCs, prepared using different dicarbonyl compounds as CAs (Scheme 3), revealed different types of coordination of CAs at Zn²⁺ ion [52]. These types of coordination were detected using FT-IR spectroscopy; clear evidence of difference in Zn²⁺ bonding with α-, β-, and γ-diketones and β-ketoesters was visible based on analysis of expanded X-ray photoelectron spectra (XPS) of DMCCs (Figure 4). The Zn 2p_1/2 and Zn 2p_3/2 spin-orbital peaks are usually ascribed to two singlets at 1044–1046 eV and 1021–1023 eV. The XPS spectra of Zn 2p in the case of DMC-AA, DMC-MAA, DMC-EAA, and DMC-TBAA (see Scheme 3) exhibited two doublets of Zn 2p_1/2 and Zn 2p_3/2. The splitting of spin-orbital peaks resulted from Zn^…O coordination. The XPS spectra clearly indicated that β-dicarbonyl CAs coordinate to the Zn²⁺ in either a keto or enol fashion. For DMC-34HD and DMC-25HD (see Scheme 3), the XPS spectra of the Zn 2p_1/2 and Zn 2p_3/2 spin-orbits were deconvoluted into two singlets (1022.9/1046.1 eV and 1023.4/1046.3 eV, respectively), indicating the presence of the only Zn^…O coordination mode. These experimental observations were confirmed by the results of the DFT modeling of CA’s coordination on the surface of Zn/Co DMCC (CASTEP module in BIOVIA Materials Studio 5.0 software package, Perdew–Burke–Ernzerhof (PBE) functional).

Although the structures of the vast majority of Zn/Co DMCCs, both laboratory and industrially implemented, were not clearly confirmed by XRD analysis, XRD studies were successful for some complexes. The structure of the previously mentioned Zn₂[Co(CN)₆]OH·4H₂O [19] served as the basis for subsequent theoretical investigations (see below). The complex Zn₂[Co(CN)₆](OAc)·4H₂O, crystallized in the monoclinic space group P2₁/m, had a layered structure (Figure 5); the asymmetric unit consisted of two symmetry independent Zn atoms and one Co atom. Zn1 was tetrahedrally coordinated to three N atoms and one O atom of the OAc⁻. Zn2 had an octahedral coordination consisting of three N atoms, one O atom from an OAc⁻ group, and two water molecules [60].

The structure of [(MeOCMe₂O)Zn⁺][Zn²⁺][Co(CN)₆³⁻] was also clearly established by XRD methods [62]. The complex crystallized in hexagonal space group P6₃/mmc, the overall structure represented a stack of layers, with each layer characterized by the composition of [(MeOCMe₂O)Zn⁺][Zn²⁺][Co(CN)₆³⁻] (Figure 6). The interlayer distance was 8.5 Å, thereby creating channels and cavities between layers. These channels and cavities host approximately two H₂O molecules (1.75H₂O) as well as a MeOCMe₂O-fragment per [(MeOCMe₂O)Zn⁺][Zn²⁺][Co(CN)₆³⁻] unit.

It should be noted that the Zn₂[Co(CN)₆]X structural motif was used and can still be used for mechanistic analysis of chemical behavior of Zn/Co DMCCs in different catalytic processes. The first attempt of density functional theory (DFT) modeling of the active sites in Zn/Co DMCCs (PW91 exchange and correlation functional for periodic calculations, B3LYP/LANL2DZ level of theory for molecular optimizations) was made by Wojdeł et al. in 2007 [29]. The aim of this study was to select an atomic cluster adequately representing the active site of the DMC catalyst. Calculations were made for the cluster model containing six atoms of Co and twelve atoms of Zn; this model was based on unpublished XRD data obtained previously for Zn₂[Co(CN)₆]OH·4H₂O. B3LYP/LANL2DZ level of theory, used in this work, raises certain questions; the values of Bader charges at C and O atoms (1.92–1.93 and −2.45, respectively) are of interest. The next attempt of quantum-chemical modeling of the structure of Zn/Co DMCC took place in 2015: López et al. took the native DMC structure (hexagonal unit cell, in which the octahedral Co unit acts as a spacer for the tetrahedral Zn polygons) to reoptimize the cells using PBE functional [99]. The authors chose to fully replace Cl anions in the lattice by OH, OEt, and O^tBu ligands, with subsequent evaluation of the sites obtained in catalytic process.

In conclusion, it should be noted that the current industry uses different types of Zn/Co DMCCs, prepared from ZnCl₂ and K₃[Co(CN)₆] in various reaction conditions (solvent, CAs, temperature, concentration (gradients), washing and drying procedures), which typically leads to divergent morphologies and crystallinities (Figure 7) as well as catalytic characteristics [104].

2.2.2. Structure of Other DMC Complexes

Divalent metal (M) hexacyanometallates M_x[M’(CN)₆]₂) are, by far, the most studied class of cyanometallate compounds. They usually crystallize in an open microporous cubic structure (Figure 8a). When the negative charge of the hexacyanometallate ([M’(CN)₆]^b−) block and the positive charge of the divalent transition metal cation M²⁺ differ, hexacyanometallates do not meet the oxidation state sum rule. Instead, some molecular [M(CN)₆]^b− blocks are randomly vacant in the structure (Figure 8b), providing formation of open metal sites with active M²⁺ centers. Comparative study of several Prussian blue analogs M₃[M’(CN)₆]₂ (M = Fe, Co, Ni, Mn, Cu; M’ = Fe, Co) revealed proximity in structure and the presence of small and large pores [105].

The powder XRD pattern reported by Zhang et al. [39] of a Zn/Fe DMC obtained by ball milling of K₃[Fe(CN)₆] and Zn(OAc)₂ was qualitatively similar to the diffraction pattern of Zn₂[Co(CN)₆](OAc) [60], and the layered structure can be proposed for this type of Zn/Fe DMC catalyst. The layered structure of the complexes M[Ni(CN)₄] (M = Ni, Co, Fe, Mn), providing their high catalytic activity, was revealed recently by Penche et al. [106].

3. Mechanistic Insights of the Action of DMC Catalysts

3.1. DMC Catalysts in Polymerization of Oxiranes

As was showed experimentally in the earlier works [38,98], Zn₃[Co(CN)₆]₂ has low activity, as well as DMC complexes prepared without excess of ZnCl₂ even in the presence of CAs, whereas excess of ZnCl₂ endows DMCCs with high activity even in the absence of CAs. The mechanism of initiation of PO polymerization, proposed in [38,98], is presented in Scheme 5. In [107], a very similar mechanism with a replacement of both Cl⁻ ligands by ROH was presented; this mechanism was mentioned in review of Herzberger et al. [7], despite the low probability of double substitution of Cl⁻ ligands by ROH (quite apart from the fact that the presence of ZnCl₂ fragment in DMCC is doubtful, see Figure 1 and Figure 2).

In comparative experimental and theoretical studies of Zn/Co DMCCs, prepared with the use of CAs of different nature (^tBuOH and dicarbonyl compounds) [52], the mechanism of polymerization of oxiranes was studied in more detail. The authors took into account the results of the study of Prussian blue analogs [105] that revealed two types of micropores located within the DMC framework: small pores formed by the defect-free unit cells and larger pores formed by the one-third random vacancies, as required by electrical neutrality. Bearing in mind estimation of the pore size in Zn-Fe/Co DMC complexes (~6.1 Å) [108], they proposed that PO monomers are too small to pass through the pores of various sizes (apparently, due to amorphous nature and the presence of small molecules eliminated during activation, real DMCCs contain pores of larger diameter). Once the monomers adsorbed in the interstitial sites are activated, the growing polymer chains induce stress to the catalyst matrix, followed by fragmentation of the catalyst particles, thereby exposing active interstitial sites. The phase changes observed during polymerization included fine dispersion of the catalyst particles at early stage, and this process was accompanied by the loss of Zn^…CA coordination. Defragmentation the catalyst particles was confirmed experimentally by dynamic light scattering (DLS) measurements [52]. The polymerization mechanism proposed in [52] is presented in Figure 9. At the initiation stage, oxirane is coordinated onto the surface Zn sites bonded with CA molecules, with subsequent nucleophilic attack of a weakly coordinating ligand (initiator, P123, H₂O) or dissociated Cl⁻ on the coordinated oxirane to generate propagating species, along with insertion of monomer unit and proceeds until monomer is consumed completely (path A in Figure 9). Initiation via coordinative route most properly proceeds on the surface of the catalyst, where additional coordinating ligands are mainly located. Alternatively, the monomers adsorbed at internal Zn²⁺ sites undergo an activated-chain end (ACE) mechanism to afford a carbocation intermediate, followed by iterative addition of a monomer until the growing species are exposed to external ligands (path B). In the absence of an initiator, the reaction may be proceeded by both ACE and coordinative mechanism until cationic growing species are terminated completely, resulting in the formation of high-MW fractions and high PDI values. In the presence of an initiator, chain propagation via ACE is suppressed.

The effect of CA, which is a part of DMCCs, is directly related to the type and energy of Zn–O coordination: in highly active catalysts, the interaction between CA molecules and Zn²⁺ sites must be neither too weak (providing substitution of H₂O during preparation of the catalyst) nor too strong (providing substitution by oxirane monomer at the initial stage of polymerization). Therefore, the dormant Zn²⁺ sites formed by the adsorption of γ-diketones or β-ketoesters (C=O^…Zn coordination) exhibited much higher activity than those in enol form of β-diketones (C–O–Zn bonding) [52].

In the frameworks of conventional concepts of the nature of catalytic species, single electrophilic Zn ions in different ligand environment represent catalytic sites. However, in many works, devoted to the coordination–insertion mechanism of ring-opening polymerization, the role of bimetallic catalyst species was proposed and proven [109,110,111,112,113]. Based on DFT modeling results for Zn₂[Co(CN)₆]Cl [52], the distance between Zn atoms (4.4 Å) meets the possibility of bimetallic coordination–insertion mechanism for Zn/Co DMCCs. This concept was further developed in theoretical studies of PO/CO₂ copolymerization (see below).

The possibility of alternative reaction mechanisms (Figure 10) meeting the kinetics of DMC-catalyzed polymerization of PO was studied and discussed in [104]. Direct experimental evidence in favor of binuclear mechanism (coordination of PO molecule and its reaction with an external polymer chain end) were based on kinetic measurements; therefore, the catalytic action of Zn/Co DMCCs can be more readily understood by considering them as an individual “microreactors”, which brings to the fore the factors of the PO diffusion rate and poly(PO) segment length.

DFT modeling of the mechanism of PO polymerization using DMCCs was presented and discussed in a few works. In 2015, the cluster model of Zn/Co DMC complex Zn₂[Co(CN)₆]X (X = Cl, OH, OEt, O^tBu) was used for the further theoretical study of the ROP of oxirane as a model substrate [99]. The calculations (PBE functional), performed with the use of Zn–OH sites, showed the preference of the ROP on Zn sites at (001) facets. Revealing the influence of the number of O–Zn bonds (three for (100) facets and two for (001) facets) is important for further understanding of the catalytic behavior of Zn/Co DMCCs. However, the consideration of Zn–OH (or Zn–OR) species as main centers of oxirane’s polymerization seems doubtful in our opinion.

DFT modeling of the ROP of PO, catalyzed by Zn/Co DMCCs, was also carried out recently by Verma et al. on a molecular level (PW91 functional, double numerical plus polarization basis set) as a part of the study of complexes, prepared with the use of Schiff bases as CAs [114]. The structures of the key intermediates and transition states for HO(CH₂)₄OH-initiated ROP of PO, catalyzed by the complex of (E)-2-((phenylimino)methyl)phenol, are presented in Figure 11. The authors proposed a single-site reaction coordination–insertion mechanism for this process; the overall calculated activation barrier for the polymerization of PO was +95.2 kJ∙mol⁻¹ and the overall reaction was exergonic by −78.2 kJ∙mol⁻¹.

The aspects of polymerization kinetics are crucial to industrial production of polyols. As applied to DMC-catalyzed polymerization of oxiranes, a kinetic model was proposed recently by Jupke et al. [115,116]. This model covers the characteristic “catch-up” kinetics (when DMCCs prefer to grow short chain polyols while long chain polyols stay unreacted), favoring the growth of relatively small macromolecules, and undesirable high molecular weight tailing. This model meets experimental results, observed in (semi-)batch operations during production of PO-based polyols, and was used in recent development of the process of polyol production by the use of a semi-batch reactor with external loop for continuous PO feed [117].

3.2. DMC Catalysts in Copolymerization of Oxiranes with CO₂

The mechanism of copolymerization of PO and CO₂ is not yet clear. The microstructure of copolymers clearly indicates the absence of anhydride fragments, since insertion of CO₂ at Zn–OC(O)R is thermodynamically forbidden [118]. The growth of a copolymer chain includes subsequent coordination and insertion of oxirane and CO₂ molecules. The degree of introduction of CO₂ F_CO2, determined as a fraction of carbonate fragments relative to the total number of ether and carbonate units, is an important characteristic of PEC; for polycarbonates (PC) F_CO2 = 100%.

Most researchers agree that PO/CO₂ copolymerization proceeds on single metal center (Zn atom for Zn/Co DMCCs) and is influenced by ligand environment of the Zr²⁺, reaction media and conditions [6]. Important mechanistic patterns were revealed when studying Zn/Co DMC-catalyzed copolymerization of cyclohexene oxide and CO₂ [119]. The authors showed that Zn–OH fragment is the real active site for initiating the copolymerization. Importantly, the introducing of ^tBuOH into the reaction mixture suppressed the formation of the oligoether fragments, ultimately leading to alternating polymer (F_CO2 = 97%).

DMC catalysts with enhanced Lewis acidity (e.g., TiO₂-supported) demonstrate high selectivity of copolymerization with low formation of cyclic carbonates. According to the authors of [75], Lewis acidic character of the TiO₂ support leads to stronger coordination of the PO molecules at the Zn centers or provides additional coordination sites for the PO molecules in the vicinity of the Zn centers. A higher concentration of PO on the surface will enhance the probability of transferring the terminal carboxylate group (pathway α → β in Figure 12) or terminal RO⁻ group to a neighboring PO molecule (β). On the contrary, the probability of dissociation of a terminal group, followed by nucleophilic attack of the alcoholate end group at a vicinal carbonate group with a formation of cyclic carbonate, is reduced (γ in Figure 12).

The difference in catalytic behavior between well-defined Prussian blue analog Zn₃[Co(CN)₆]₂∙5H₂O and amorphous Zn/Co DMC complex, prepared using Me₃SiOH and P-123 as CAs, was clearly demonstrated by Kim et al. by the example of the reaction of glycidol with CO₂: in the former case, cyclic carbonate was formed, and in the latter case, copolymerization was observed [66]. Formation of glycerol carbonate was explained using DFT modeling (PBE generalized gradient approximation functional, double numerical basis set with polarization) on the assumption of the participation of surface Zn–OH species in the catalytic cycle; the calculated activation barrier for this reaction (Figure 13a) amounted to ~10 kcal∙mol⁻¹ (G scale). More acidic and less sterically hindered Zn sites in amorphous Zn/Co DMCC initiate polymerization (Figure 13b).

It also should be noted here that the concept of binuclear mechanism of the action of Zn/Co DMC catalysts was also applied to PO/CO₂ copolymerization. Based on results of previous studies [99], Stahl and Luinstra proposed two bimetallic reaction pathways, distinguishing by the nature of catalytic sites (non-basic and basic, see Figure 14) [120] and including only low activation energy transformations. The reaction sequence may be formulated in form of an external attack of a hydroxyl chain end, proton transfer, and liberation of a coordination site for PO activation (Figure 14, N1–N4) with assistance of Cl ligand (N3). The incorporation of CO₂ occurs via insertion into a Zn–OR bond (N5). Subsequent protonation would lead to unstable alkyl carbonic acid (N6) that decarboxylates (non-productive cycle). Interruption of this cycle is achieved when the intermediate carboxylate is not protonated but reacts with activated PO to regenerate Zn–OR (Figure 14, N7–9 or N7, N10). This catalytic scheme was proposed for DMC-Cl systems with zero response of PEC composition on CO₂ concentration in the feed. When using DMC-NO₂ catalysts, prepared during studies, the CO₂ incorporation was positive dependent on CO₂ concentration, which found explanation in the frameworks of the mechanism presented in Figure 15. After exchange with hydroxyl compounds in the solution, anionic chain ends (B1), formed on the surface, may insert CO₂; the CO₂ insertion rate becomes dependent on the CO₂ concentration in the feed, and this process competes with propoxylation (B5–B7).

3.3. Mechanism of Transesterification of Cyclic Carbonates

Transesterification of cyclic carbonates (see Section 4.6) is not the most important field of DMCC application; however, the mechanism of this reaction deserves a mention in view of the bimetallic concept mentioned earlier. As was shown in [121], methanolysis of propylene carbonate, catalyzed by Mn/Fe DMC complex, can be described by the following mechanistic steps (Scheme 6): (1) activation of MeOH by electrophilic Mn center, (2) interaction of propylene carbonate with activated MeOH to form intermediate II, (3) reaction of the intermediate II with activated MeOH (second Mn atom) to form intermediate III, (4) decomposition of intermediate III with a formation of (MeO)₂C=O and propylene glycol, completing the catalytic cycle.

4. Actual Processes and Products

4.1. Preparation of Polyols

The main industrial application of DMCCs is their use in production of high molecular weight polyalkylene ether polyols; propylene oxide is the most used monomer (Scheme 7a) [53].

Since 1966 [17], polymerization of PO and other substituted oxiranes has been intensively studied and used for the production of high-quality polyols that are highly in demand in the production of polyurethanes [20,122,123,124]. In 2005, Ionescu completed a detailed review of the use of DMCCs in the synthesis of polyols [125]. However, in the years since, significant progress in this area has been made, and catalyst concentrations of as little as 15–50 ppm are sufficient for industrial production of polyols [7].

4.1.1. Zn/Co DMCCs in the Synthesis of Polyols

Preparation of polyols using Zn/Co DMCCs is complicated by the impossibility of using low-MW initiators such as propylene glycol and glycerol. It is assumed that 1,2-diols and glycerol form stable chelate complexes, thereby inhibiting coordination of monomers [24] (Scheme 7b). As a consequence, polymerization of PO with the use of Zn/Co DMCCs is usually initiated by oligomeric polyols, prepared by KOH-catalyzed condensation of PO [10]. Besides these polyols, low-MW paraformaldehyde [126], castor oil [127], oxalic acid [128], 1,3,5-benzenetricarboxylic acid [129], and 1,2,4,5-benzenetetracarboxylic acid [130] were used as initiators. Note that Zn/Co DMCC, prepared by the reaction of H₃[Co(CN)₆] with Zn(II) 2-ethylhexanoate, was active in polymerization of PO, initiated by propane-1,2-diol [61]. In 2024, Verma et al. reported the results of kinetic experiments of PO polymerization, catalyzed by Zn/Co DMC complex, prepared with the use of EDTA as a CA; for this type of catalyst, ethylene glycol showed no chelating effect and acted as an initiator and chain transfer agent [131].

Another critical concern in DMC-catalyzed polyol production is the formation of high-MW impurities [99,117,120,132] and unsaturations [53], attributed to action of Zn–OH initiators, initially presented in DMCC or formed as a result of dehydration of ^tBuOH (Scheme 7c) and conversion of methyloxirane to allyl alcohol or termination of the polyether chain with a formation of PPO containing allyl end-group (Scheme 7d).

Polymerization of PO, catalyzed by Zn/Co DMCCs, is usually conducted in bulk at elevated temperatures (e.g., 120 °C, polymerization time 4–8 h [133]); ¹³C NMR analysis of poly(propylene oxide) (PPO) showed high regioselectivity of polymerization (head-to tail sequence). PPO, obtained with the use of Zn/Co DMCCs, had narrow MWD and low content of unsaturations. In early works [38], studies of the influence of DMCCs composition and morphology on their characteristics also revealed that the amorphous catalyst prepared from ZnCl₂ and ^tBuOH had the highest activity. The addition of minimal amounts of CaCl₂ to DMC formulation increased its catalytic activity [107], but the reasons for this phenomenon have not been explained.

The synthesis of Zn/Co DMCCs with the use of ^tBuOH as a CA requires high consumption of ^tBuOH. The use of lactates in combination with polyols results in catalysts that allow obtaining of high-quality polymers (M_n = 4.1–4.9 kDa, Ð_m = 1.08–1.25) with unsaturation lower than 0.0034 meq∙g⁻¹ [107]. Replacement of ^tBuOH by EDTA at the stage of preparation of Zn/Co DMCC resulted in obtaining of a catalyst with similar morphology, but low TOF = 9625 h⁻¹ and unsaturation level of 0.009 meq∙g⁻¹ [67]. Promising results in the development of efficient catalysts for the synthesis of polyols were obtained recently by Verma et al.: when using Schiff bases instead of ^tBuOH with PEG as a co-CA, obtained Zn/Co DMCCs showed very high catalytic activities with a short induction period, outperforming conventional ^tBuOH-based catalysts on these parameters [114]. The use of cyclic carbonates instead of ^tBuOH at the stage of the preparation of Zn/Co DMCCs resulted in a twofold increase of TOF from 212 to 467 min⁻¹ (115 °C, 85 ppm of the catalyst) even though much lower amounts of CA and ZnCl₂ were used during preparation of the catalyst [58]. A similar increase of the catalytic activity was detected for Zn/Co DMCCs, prepared with the use of dialkyl phosphites or trialkyl phosphates as Cas; e.g., (EtO)₂P(O)H-based catalyst showed TOF = 725 min⁻¹ at 115 °C [59].

Relatively recently [62], a number of well-defined Zn/Co DMCCs were tested in comparison with the benchmark DMC (Zn/Co ratio ~2:1) currently used in the industry. The experiments were conducted with a continuous feed of PO, making it difficult to compare inherent activities of the catalyst’ samples; however, comparison with benchmark DMC allows estimation of this parameter. As can be seen in

Table 1. PO polymerization results of the prepared DMC catalysts [62] (for catalyst legends, see Scheme 3) ¹.

Entry	Catalyst	Init. Time, min	Cat. Longevity, min	Polymer Yield, g	TOF, ([PO]∙ [Zn]−1∙h−1)	Unsatur. Level 2	Mn, kDa (ÐM)
1	Benchmark DMC	1	29	27.5	132,000	0.004	5.69 (4.88)
2	[DMC-Cl][MeOCMe2OH]	heating	30	27.9	120,000	0.024	6.05 (1.18)
3	[DMC-Cl][tBuOH]	4	29	28.4	126,000	0.022	6.17 (1.16)
4	DMC-OCMe2OMe	3	25	23.0	119,000	0.004	2.62 (3.70)
5	DMC-(OCMe2OMe/Cl)(0.25/0.75)	2	32	31.5	127,000	0.018	5.33 (1.40)
6	DMC-(OCMe2OMe/Cl)(0.5/0.5)	3	30	29.9	129,000	0.014	6.11 (3.57)
7	DMC-(OCMe2OMe/Cl)(0.75/0.25)	2	27	26.2	125,000	0.006	5.44 (2.00)
8	DMC-OtBu	heating	27	26.6	127,000	0.004	6.07 (7.17)
9	DMC-(OtBu/Cl)(0.5/0.5)	7	31	28.7	120,000	0.005	6.40 (6.44)
10	DMC-OAc (60 °C, evacuation)	2	29	30.0	134,000	0.016	6.91 (2.53)
11	Zn2[Co(CN)6](OAc)·4H2O	1	28	27.3	126,000	0.014	6.22 (11)
12	DMC-(OAc/OtBu)(0.95/0.08)	0	30	31.9	137,000	0.013	7.15 (1.75)
13	DMC-(OAc/OtBu)(0.68/0.31)	0	29	29.6	132,000	0.010	7.06 (4.02)
14	DMC-(OAc/OtBu)(0.42/0.57)	4	28	28.2	130,000	0.007	6.48 (10.7)
15	DMC-(OAc/OtBu)(0.31/0.71)	heating	24	24.9	134,000	0.005	6.10 (9.12)
16	DMC-(OAc/OtBu/Cl)(0.68/0.38/0.06)	0	29	29.9	133,000	0.010	6.94 (4.17)
17	DMC-(OAc/OtBu/Cl)(0.74/0.24/0.10)	1	29	30.1	134,000	0.008	6.30 (3.08)
18	DMC-(OAc/OtBu/Cl)(0.59/0.38/0.15)	0	29	30.0	134,000	0.007	6.56 (3.32)
19	DMC-(OAc/OtBu/Cl)(0.41/0.43/0.28)	1	28	29.2	134,000	0.005	6.28 (3.73)
20	DMC-(OAc/OtBu/Cl)(0.34/0.39/0.39)	2	31	31.0	129,000	0.006	6.58 (3.42)

¹ Polymerization conditions: DMC (4.0 μmol Co, 1.6−1.9 mg), PPG starter (M_n = 425 Da, 4.25 g), heating to 125 °C, feeding 5.0 mL PO for 4 min for initiation, waiting until almost full consumption of PO (initiation time), continuous feed of PO (1.25 mL∙min⁻¹) at 130−145 °C with spontaneous temperature control. ² Calculated based on ¹H NMR spectra.

, even in the absence of Cl in its composition (Table 1, Entry 12), the catalyst of the formula DMC-(OAc/OtBu)(0.95/0.08) displayed the yield surpassing that of the benchmark DMC. Following closely were the DMC-(OAc/OtBu/Cl) series (29–31 g yields) and DMC-Cl (28 g yield). The initiation time was quite short (several minutes) in most cases, except [DMC-Cl][MeOCMe₂OH], DMC-OtBu, and DMC-(OAc/OtBu)(0.31/0.71), which required external heating for initiation. The polyols generated with DMC-Cl exhibited high unsaturation levels (0.022–0.024 meq∙g⁻¹), whereas those produced by DMC-OCMe₂OMe and DMC-OtBu showed very low unsaturation levels, matching that of the benchmark DMC (0.004 meq∙g⁻¹). Zn/Co DMC complexes, prepared with the use of different dicarbonyl compounds, showed broad spectrum of catalytic properties; the highest productivity (TOF = 569 min⁻¹ at 130 °C) was demonstrated by the complex prepared using diethyl malonate as a CA, and the obtained polymer had M_n = 2.9 kDa, Ð_m = 1.02 and unsaturation of 0.0017 meq∙g⁻¹ [57].

To increase catalyst activity, stability and reusability, supported Zn/Co DMCCs were prepared in recent years. In particular, MCM-41@Zn/Co DMCC (0.05 wt%) catalyzed polymerization of PO with 90.6% conversion, not inferior to conventional DMCCs, but showed good reusability (at least 7 cycles) [72]. To access high-MW polyols, Kim et al. proposed a sequential approach, based on conducting of a series of successive stages, affording PEs with M_n up to 43.6 kDa [134].

Besides PO, other oxiranes were studied in homopolymerizations catalyzed by Zn/Co DMC complexes. For example, epichlorohydrin was polymerized using tri-(2-hydroxyethyl) isocyanurate as an initiator, and the reactivity of the monomer was comparable with reactivity of PO [135]. Regioselective ROP of enantiomerically pure epichlorohydrin on Zn/Co DMCC was not accompanied by epimerization; a solid polymer was obtained in contrast with the liquid product of the homopolymerization of rac-epichlorohydrin (T_g = −38 °C) [136].

Preparation of hyperbranched polyols via copolymerization of PO and glycidol was studied by Gu and Dong [137]. During later studies, Kim et al. showed that Zn/Co DMC complexes in combination with diethyl malonate or ethyl acetoacetate efficiently catalyze homopolymerization of glycidol without the use of an initiator; degrees of branching (DB) were determined by the type of monomer addition method: DB = 0.25–0.27 for solvent-free batch process and DB = 0.51–0.57 for semi-batch polymerization [138]. To incorporate polar structural units to poly(PO), Frey et al. proposed the use of glycidyl methyl ether (GME) as a comonomer; commercial DMC catalyst from BASF was applied in their studies [139]. Copolymers with M_n = 1.9–4.5 kDa, Ð_M < 1.29 and up to 45 mol% GME content were obtained. It is noteworthy that copolymers had a random microstructure (r_PO = 1.40 ± 0.01, r_GME = 0.71 ± 0.01), which qualitatively distinguishes GME from other substituted oxiranes.

In 2024, a zeolite-supported layered Zn/Co DMC was used as a catalyst for polymerization of butylene oxide [76]. The high catalyst’s performance was attributed to the high dispersion of DMC and the presence of additional Lewis acid sites on the support surface. Products with M_n = 1.4–4.5 kDa, Ð_M = 1.08–1.53 were obtained. Poly(butylene oxides) were characterized by viscosity index of up to 242 and relatively low pour point values (as low as −45 °C) which made it possible to consider these materials as an additives to mineral lubricating oils.

4.1.2. Other DMCCs in the Synthesis of Polyols

As was shown in [79], polymerization of PO in the presence of Zn/Fe DMCCs, prepared using Na₂EDTA as a CA, can be initiated by ethylene glycol. Under optimized conditions (350 ppm of the catalyst, 105 °C, 5 bar, 24 h), PPO with Ð_M = 1.29 and OH value of 4.9 mg_KOH∙g⁻¹ was obtained; during optimization, a wide spectrum of polymers (M_w = 0.16–19.67 kDa, Ð_M = 1.05–3.01) were synthesized.

Zn/Ni DMCCs, prepared in the absence of CA by subsequent reaction of NiCl₂ with KCN and Zn(OAc)₂ ([NiCl₂]/[KCN]/[Zn(OAc)₂] = 1:6:6), demonstrated high catalytic activity in polymerization of PO and other oxiranes, but were inferior to Zn/Co DMCCs in activity [89]. Co/M DMCCs, synthesized from CoCl₂ and K₂[Ni(CN)₄], K₄[Fe(CN)₆], K₃[Fe(CN)₆], or K₃[Co(CN)₆] in the presence of ^tBuOH and PTMEG, were less active in comparison with the Zn/Co DMC complexes; only Co/Ni DMCC approached the benchmark catalyst in productivity [85].

4.1.3. Post-Functionalization of PEs

Low-MW polyols, obtained by ROP of PO, are usually intended for the production of polyurethanes. However, these materials, obtained using DMCCs, contain a low degree of primary hydroxyl groups having higher reactivity at the stage of the reaction with isocyanates. The commonly use procedure of the introduction of –CH₂OH fragments to PE microstructure is a base-catalyzed ethylene oxide capping [140]. To simplify the preparation of –CH₂OH-terminated polyols, Raghuraman et al. proposed the use of tris(pentafluorophenyl)borane at the final stage of ROP of PO (Scheme 8) and co-ROP of PO and butylene oxide; the content of more than 65% of primary hydroxyl end-groups was achieved [141], and the loading of ~150 ppm B(C₆F₅)₃ turned out to be sufficient.

4.2. Copolymerization of Oxiranes with CO₂ and Other C1 Comonomers

Copolymerization of oxiranes with CO₂ may result in the formation of poly(ether-carbonates) and polycarbonates (PEC and PC, respectively); the latter copolymers represent macromolecules with alternating sequence of comonomer units [8,15] (Scheme 9). It is believed that DMCCs usually catalyze formation of PEC; however, in some cases, copolymers, obtained in the presence of DMCCs, are approaching the composition of PCs. The products of DMCC-catalyzed copolymerization of PO and CO₂ represent polyols with carbonate inserts that are also of interest to polyurethane production [142]. Cyclic carbonates are side products of this reaction. Given that cyclic carbonates are demanded fine chemicals, this process appears to be waste-free and environmentally friendly, which explains continued interest of the researchers to DMC-catalyzed copolymerization of oxiranes and CO₂. F_CO2 is an essential property of PO/CO₂ copolymers: as was shown in [143], polyols with high carbonate content provide higher thermo-oxidative stability of polyurethanes with reduced generation of volatile organic compounds (e.g., acetaldehyde) during decomposition. Another substantial factor is a structure of initiator/chain transfer agent/chain extender used during preparation of PO/CO₂ copolymers [144]; in this particular case, fine tuning of these chemicals allowed to select copolymer suitable for the production of polyurethanes compatible with anti-corrosive polyanilines.

PO is a most commonly used monomer, and splitting on subsections below was made in view of this fact. The products of copolymerization of other oxiranes with CO₂ turned out to be less attractive for their use as a material of the new millennium with a few exceptions (see below). Other C1 comonomers (COS, CS₂) can also react with oxiranes in the presence of DMCCs, and these processes will also be considered in this Section.

4.2.1. Zn/Co DMCCs in Copolymerization of Propylene Oxide and CO₂

In comparison with ROP of PO, copolymerization of PO with CO₂ proceeds much more slowly. The process is limited by the same factors as the constraints and challenges of PO homopolymerization except formation of high-MW polymer fraction. In early study of Gao et al. [145], PECs with M_n = 1.8–6.4 kDa, Ð_M = 1.14–1.83 and F_CO2 = 15.3–62.5% were obtained, and productivity of Zn/Co DMCC (prepared by standard method from K₃[Co(CN)₆], ZnCl₂ and ^tBuOH) was up to 10 kg∙g_cat⁻¹. The formation of propylene carbonate was not more than 8%. PO oligomers were used as initiators in these experiments. Zn/Co DMCCs, prepared using ZnX₂ (X = F, Cl, Br, I), ^tBuOH, and poly(tetramethylene ether glycol), were studied in PO/CO₂ copolymerization [27]. ZnF₂-based catalyst showed negligible activity, whereas other ZnX₂-based catalysts demonstrated comparable productivities, ~510 g∙g_cat⁻¹ (50 °C, 9.6 bar CO₂, 24 h), F_CO2 were 22, 36, and 32% for X = Cl, Br, and I, respectively. When using polyvinylpyrrolidone as a co-CA during preparation of Zn/Co DMCC with standard co-precipitation in the presence of ^tBuOH, the modified catalyst exhibited higher catalytic activity than the conventional DMCCs, with a nearly 10% increase in PO conversion. The most critical breakthrough was a reduction of the formation of propylene carbonate to ~5 wt% [146]. It was also shown that ^tBuOH can be substituted by pluronic P-123 at the stage of preparation of Zn/Co DMCCs, but careful calcination is critical to providing reduced crystallinity and enhanced formation of the active monoclinic phase [147].

Conventional Zn/Co DMCCs usually provide moderate carbonate content (up to 40 mol%), but the use of [ZnCl]₂[HCo(CN)₆]∙MeOH resulted in copolymers with F_CO2 ~ 60% [44], introducing of 1,10-decanediol as chain transfer agent provided the formation of low-MW PECs (M_n ~ 2 kDa) suitable for polyurethane production. The later study of Zn/Co DMCC-catalyzed PO/CO₂ copolymerization revealed the preference of the formation of a copolymer containing 3–4 ether units, thereby confirming the mechanism of PO/CO₂ copolymerization initiated by moderately acidic Lewis centers like Zn in Zn/Co catalysts [148].

Besides PO-based polyols, bisphenol A [149], 1,1,1-tris(4-hydroxyphenyl)ethane [150], oxalic acid [128], sebacic acid [151], low-MW paraformaldehyde [126], and polycarboxylic aromatic acids [129,130,152] were used as initiators. It was also shown that dicarboxylic acids may act as a chain transfer agents, thereby controlling molecular weight characteristics of PEC [151]. The nature of the initiator allows it to affect the architecture of the macromolecules (the number of –OH end groups), which is essential for the further use of PEC in polyurethane production.

A comparative study of a number of well-defined Zn/Co DMCCs and industrial DMC [62] was conducted with continuous feed of PO and CO₂. As can be seen in

Table 2. PO/CO₂ copolymerization results of the prepared DMC catalysts [62] (for catalyst legends, see Scheme 3) ¹.

Entry	Catalyst	Init. Time, min	Polymer Yield, g	TOF, [PO]∙ [Zn]−1∙h−1; [CO2]∙ [Zn]−1∙h−1	Cyclic Carbonate, wt% 2	FCO2, mol% 3	Mn, kDa (ÐM)
1	Benchmark DMC	36	25.4	19,900; 2980	2.5	15	5760 (3.34)
2	[DMC-Cl][MeOCMe2OH]	54	5.6	4070; 1060	25	26	1420 (3.04)
3	[DMC-Cl][tBuOH]	25	3.1	2280; 547	16.2	24	1120 (1.78)
4	DMC-OCMe2OMe	191	3.4	2640; 422	9.2	16	1270 (1.55)
5	DMC-OtBu	111	25.9	20,200; 3030	1.3	15	6250 (4.04)
6	DMC-(OCMe2OMe/Cl)(0.5/0.5)	42	6.7	4930; 1180	16	24	1630 (2.07)
7	DMC-(OtBu/Cl)(0.5/0.5)	37	21.7	16,500; 3130	2.4	19	5450 (8.81)
8	DMC-OAc (60 °C, evacuation)	39	18.7	14,800; 1920	4.9	13	3960 (1.76)
9	Zn2[Co(CN)6](OAc)·4H2O	45	6.8	4850; 1410	7.4	29	1960 (1.73)
10	DMC-(OAc/OtBu)(0.95/0.08)	13	26.4	18,500; 5910	3.5	32	5420 (1.43)
11	DMC-(OAc/OtBu)(0.68/0.31)	15	28.4	20,800; 5190	4.0	25	3380 (5.39)
12	DMC-(OAc/OtBu)(0.42/0.57)	31	19.7	14,400; 3600	3.4	25	5970 (4.20)
13	DMC-(OAc/OtBu)(0.31/0.71)	49	19.3	11,500; 3040	2.4	21	4810 (9.32)
14	DMC-(OAc/OtBu/Cl)(0.68/0.38/0.06)	21	24.0	17,800; 4090	2.7	23	5510 (5.41)
15	DMC-(OAc/OtBu/Cl)(0.74/0.24/0.10)	17	25.8	19,000; 4560	1.1	24	5790 (3.33)
16	DMC-(OAc/OtBu/Cl)(0.59/0.38/0.15)	15	26.3	19,500; 4480	1.8	23	5830 (2.63)
17	DMC-(OAc/OtBu/Cl)(0.41/0.43/0.28)	22	25.6	19,100; 4200	2.8	22	6070 (2.71)
18	DMC-(OAc/OtBu/Cl)(0.34/0.39/0.39)	14	23.0	17,500; 3320	3.2	19	5510 (3.28)
19 4	DMC-(OAc/OtBu)(0.95/0.08)	20	18.2	12,000; 5040	3.7	42	4290 (1.43)
20 5	DMC-(OAc/OtBu)(0.95/0.08)	7	22.2	15,400; 5100	3.5	33	4960 (1.86)
21 6	DMC-(OAc/OtBu)(0.95/0.08)	13	25.3	34,100; 10,900	3.4	32	5420 (1.43)
22 7	DMC-(OAc/OtBu)(0.95/0.08)	6	5.4	3480; 1600	3.2	46	35.50 (7.17)

¹ Polymerization conditions: DMC (11 μmol-Co, 4.5–5.2 mg), PPG starter (M_n = 425 Da, 4.25 g), heating to 110 °C, charging CO₂ to 30 bar, feeding 5.0 mL PO for 4 min for initiation, waiting until pressure drops from 32 to 30 bar (initiation time), continuous feed of PO at 0.50 mL∙min⁻¹ for 54 min with temperature control using external heating, charging CO₂ to 30 bar whenever pressure drops to 25 bar. ² (Propylene carbonate mass)/(polymer mass + propylene carbonate mass) determined on ¹H NMR spectra. ³ (Carbonate linkage)/(ether linkage + carbonate linkage) in polymers excluding cyclic carbonates determined on ¹H NMR spectra ⁴ Polymerization temperature of 100 °C. ⁵ Pressure of 40–35 bar. ⁶ Two times higher PO feeding rate (1.0 mL/min) and half-times shortened reaction time (28 min). ⁷ Copolymerization was performed by replacing the starter with THF (4.25 g) at 100 °C with the PO feed rate of 1.0 mL/min for 28 min.

in comparison with Table 1 (PO homopolymerization), the catalytic activity decreased significantly under CO₂ pressure, and a higher amount of catalyst was needed (11 vs. 4.0 μmol-Co) with a lower PO feed rate (0.50 vs. 1.25 mL∙min⁻¹). The catalysts containing (tBuO)Zn⁺ or (AcO)Zn⁺ units, including the benchmark DMC, showed high activities, yielding 18–28 g of polyols. A dramatic productivity difference was observed between DMC-OAc and the structurally similar Zn₂[Co(CN)6](OAc)·4H₂O (19 vs. 6.8 g, respectively). DMC-(OAc/OtBu)(0.95/0.08) and DMC-(OAc/OtBu)(0.68/0.31) exhibited the highest yield (26–28 g) with short initiation time (15 min), proving superior to the benchmark DMC (25 g yield; initiation time 36 min). In terms of the dispersity of the resulting polymer, DMC-(OAc/OtBu)(0.95/0.08) also demonstrated a fairly narrow unimodal GPC curve with the lowest Ð_M = 1.43. Formation of propylene carbonate is a critical concern in PO/CO₂ copolymerization. Catalysts with poor activities, such as DMC-Cl, DMC-OCMe₂OMe, and DMC-(OCMe₂OMe/Cl)(0.5/0.5) initiated significant cyclic carbonate formation (9.2–25 mol%). The catalysts with good activities produced small amounts of cyclic carbonate (<5 mol%). Among these, DMC-OtBu, DMC-(OAc/OtBu/Cl)(0.59/0.38/0.15), and DMC-(OAc/OtBu/Cl)(0.74/0.24/0.10) performed with carbonate formation below 2 wt%. It is very interesting and important that low unsaturation levels in PO polymerization correlate with low cyclic carbonate formation in PO/CO₂ copolymerization. Both side processes can be catalyzed by homogeneous Zn species [153]; therefore, the possibility of leaching of such homogeneous species could be minimized during preparation of the DMCCs.

Interesting results were obtained in PO/CO₂ copolymerization, catalyzed by Zn/Co DMCC, prepared using ball milling, in the presence of γ-Al₂O₃ as a co-catalyst [154]. Polymerization proceeded with higher rates, and the thermal stability of copolymers was greatly improved, up to 379 °C. Another way of improving the yield and molecular weight characteristics of PECs was the use of hybrid catalyst comprising Zn/Co DMCC and Zn(II) glutharate in 1:10 molar ratio by Zn [155]; M_n up to 58 kDa, F_CO2 = 97.7%, and productivity of 508 g∙g_cat⁻¹ were achieved.

Concluding this Section, it should be noted that the properties of copolymers, obtained with the use of Zn/Co DMCCs, may be largely affected by the type and quantity of the initiator used (in other words, relatively large amounts of specific initiators can be considered as a valuable components). For example, the use of bisphenol A as an initiator and chain transfer agent allows copolymers to be obtained, which can be used in further synthesis of polyurethanes with improved flame-retardant properties [149].

4.2.2. Other DMCCs in Copolymerization of Propylene Oxide and CO₂

Zn/Fe DMCCs, prepared by ball milling, were used in copolymerization of PO with CO₂; however, high yields of cyclic carbonates were detected [80,81]. The higher CO₂ incorporation during PO/CO₂ copolymerization for Zn/Ni DMCCs in comparison with Zn/Co catalysts was revealed by Qi et al. in 2007 [28]. Nanosized Zn/Ni and Co/Ni DMCCs, synthesized by ball milling, in alternating copolymerization of PO with CO₂ showed high catalytic activity; obtained after 24 h (40 bar, 60 °C), PEC had M_n up to 10.3 kDa, Ð_M = 1.45 and 83.5 mol% of carbonate linkages (Zn/Ni), whereas Co/Ni system catalyzed formation of PPC with M_n = 8.48 kDa, Ð_M = 1.44 and F_CO2 = 73.7% [156]. Zn/Ni DMCC, prepared by co-precipitation using K₂[Ni(CN)₄], ZnCl₂, and ^tBuOH, exhibited productivity ~500 g∙g_cat⁻¹, with F_CO2 up to 60% [90].

Zn/Cr DMCC (1 wt%) at 70 °C and 40 bar catalyzed formation of PEC with M_n = 68.6 kDa and Ð_M = 1.68; F_CO2 = 38.5% [93]. DMC complexes of the formula Co[M(CN)₄] showed low (M = Pd, Pt) or moderate (M = Ni) activities in PO/CO₂ copolymerization: at [PO]/[Co] ratio of 2530:1 at 130 °C and 54 bar of CO₂ the maximum TOF was 1860 h⁻¹; for copolymer with M_n = 74.3 kDa and Ð_M = 3.1, the F_CO2 value was 20% [92]. Although the activity of Co[Ni(CN)₄] was significantly lower than the Zn/Co DMCCs, the advantage of this system was the absence of propylene carbonate in the reaction products.

Hexane-1,6-diol initiated PO/CO₂ copolymerization, catalyzed by Co/Ni DMCCs, under optimized conditions (120 °C, 43 bar of CO₂) resulted in PECs (1.58 kg∙g_cat⁻¹∙h⁻¹) with M_n = 2.6 kDa, Ð_M = 2.5, and F_CO2 = 25% [91]. Ni₃[Co(CN)₆]₂, Co₃[Co(CN)₆]₂, Fe₃[Co(CN)₆]₂, Ni₃[Fe(CN)₆]₂, and Co₃[Fe(CN)₆]₂ showed moderate activities in PO/CO₂ copolymerization; the obtained PECs had F_CO2 = 16–33%, M_w = 6–85.4 kDa, and Ð_M = 4.1–15.8 [86].

Relatively recently, hexacyanoferrate(II) complexes Ni₂[Fe(CN)₆], Co₂[Fe(CN)₆], KFe[Fe(CN)₆], and Zn₂[Fe(CN)₆], synthesized by co-precipitation in the presence of ^tBuOH, were studied in PO/CO₂ copolymerization (90 °C, 20 bar CO₂, 24 h) [87]. PECs with M_w = 3.4–20.2 kDa, Ð_M = 4.0–5.4, and CO₂ content of 9.3–18.1 wt% were obtained; the formation of propylene carbonate (1.4–19.8 wt%) was also detected. The TON values (in mol_PO per mol_M(II)) were 84–223; under the same conditions, Zn/Fe DMCC demonstrated TON of 1279. The

Table 3. Comparison of the performance of the synthesized transition metal hexacyanoferrate(II) complexes with the performance of other hexacyanometallate catalysts [87] ¹.

Catalyst	TON 2	TOF 3	FCU 4, mol%	WPC 5, wt%	SCO2 6, %	SPO 6, %	RPEC 7, %	Mw, kDa	ÐM
Ni2[Fe(CN)6]	93	4	14.9	1.4	94.2	99.1	92.4	3.6	5.4
Co2[Fe(CN)6]	223	9	29.3	5.4	88.2	96.2	67.0	20.2	4.0
KFe[Fe(CN)6]	170	7	13.5	19.8	46.6	86.6	85.3	14.0	5.1
Zn2[Fe(CN)6]	178	7	19.8	7.5	76.7	95.0	83.8	3.4	4.4
Ni3[Co(CN)6]2	86	4	22.3	0.4	98.9	99.7	79.5	11.8	10.5
Co3[Co(CN)6]2	544	23	20.0	4.2	87.4	97.2	71.9	68.6	4.1
Fe3[Co(CN)6]2	428	18	16.3	8.4	73.6	94.5	75.1	85.4	6.3
Ni3[Fe(CN)6]2	84	4	24.0	0.6	98.4	99.6	73.9	11.7	15.8
Co3[Fe(CN)6]2	296	12	33.5	13.3	75.4	90.2	66.0	50.0	5.9
Fe4[Fe(CN)6]3	162	7	17.3	43.1	26.2	67.2	83.3	6.0	8.4
Zn3[Co(CN)6]2	1279	53	9.7	8.9	61.9	94.4	86.6	8.3	2.5

¹ Polymerization conditions: 24 h at 90 °C and 20 bar of CO₂. Catalysts concentration = 2500 ppm, PO volume = 50 mL. ² In mol_PO∙mol_M(II)⁻¹. ³ In mol_PO∙mol_M(II)⁻¹∙h⁻¹. ⁴ Carbonate unit content, molar fraction of CO₂ in the polymer. ⁵ Weight fraction of propylene carbonate. ⁶ Selectivities of CO₂ and PO, calculated based on ¹H NMR spectral data [86]. ⁷ The poly(ether-carbonate)-to-polycarbonate linkage ratio, calculated based on ¹H NMR spectral data [157].

gives a visual representation of catalytic performance of different DMC complexes in PO/CO₂ copolymerization.

The complexes of the common formula M[Ni(CN)₄] (M = Ni, Co, Fe, Mn) catalyzed PO/CO₂ copolymerization yielding random PECs with M_n = 11.0–36.5 kDa, Ð_M = 2.5–5.0 and F_CO2 = 13.1–42.4% [106]. Under optimized conditions, the Co/Ni DMCC provided the formation of copolymer and only 0.1 wt% of propylene carbonate. When compared to Zn/Co DMCC, the Co[Ni(CN)₄], while exhibiting lower productivity and CO₂ incorporation, outperformed the benchmark complex by the lack of induction time, higher copolymer MW, and selectivity of CO₂ incorporation.

Among the latest publications on the subject, the work of Tebandeke et al. [158] also deserves attention: a composite catalyst, prepared from Zn(II) glutharate and Zn₃[Cr(CN)₆]₂ complex in 15:1 ratio, showed low productivity (48 g∙g_cat⁻¹) but high F_CO2 = 85.4%. This composite showed higher catalytic activity and selectivity compared to the individual components.

4.2.3. Copolymerization of Other Oxiranes and CO₂

Cyclohexene oxide (7-oxabicyclo [4.1.0]heptane) is a convenient substrate for the studies of catalytic copolymerization of disubstituted oxiranes with CO₂. Already in the first study on the theme [26], TON of 3300 and F_CO2 = 44% were achieved for Zn/Co DMCC, prepared with the use of ^tBuOH (90 °C, 36 bar, 2 h). The results of early studies in this field are described and discussed in our review [6], and productivities up to 8700 kg∙g_Zn⁻¹ and F_CO2 up to 97% were observed when using optimized catalysts and conditions [119]. When using microwave irradiation, for Zn/Co DMCCs TOF = 25,177 h⁻¹ was achieved [34]. Copolymers of cyclohexene oxide and CO₂ represent amorphous materials, and the T_g value depends on F_CO2; e.g., for copolymers with M_n ~ 5–10 kDa, T_g of 71.4, 81.4, and 98.6 °C corresponded to F_CO2 = 35, 63, and 65%, respectively [50].

Copolymerization of glycidol and CO₂ with a formation of highly branched polymers was efficiently catalyzed by amorphous Zn/Co DMCC, prepared by the reaction of ZnCl₂ solution, containing trimethylsilanol, with K₃[Fe(CN)₆] solution, with subsequent treatment with aq. trimethylsilanol/P-123 [66].

Copolymerization of bulky oxiranes, e.g., styrene oxide, with CO₂ is very sensitive to morphology of DMCCs. Nanolamellar Zn/Co DMCCs, prepared using ^tBuOH, THF, or MeOCH₂CH₂OH as CAs (~0.1 wt%) were studied in styrene oxide/CO₂ copolymerization at 30–90 °C and 20–50 bar of CO₂ [159]. In some experiments, highly regioselective copolymers with alternating degree > 99% were obtained; 2-benzyloxirane showed very similar behavior in copolymerization. Silica-supported Zn/Co DMCCs, prepared by sol–gel co-precipitation method [74], also showed high activity in copolymerization of styrene oxide with CO₂, but the formation of cyclic carbonate was more pronounced (up to 17.4% when copolymer with ~1:1 comonomer ratio was obtained).

Comprehensive comparative study of the effect of substituent in oxirane on oxirane/CO₂ copolymerization and properties of copolymer was conducted by Zhang et al. in 2015 for a sample of 11 oxiranes and Zn/Co DMCC [160]. Oxiranes with bulky substituents (2,2-Me₂, ^tBu, Cy, n-C₁₀H₂₁, Bn) formed copolymers with F_CO2 up to 100%. The regioselectivity of oxirane/CO₂ copolymerization was influenced by the electron induction effect: the electron-withdrawing substituents (Ph, Bn) induced regioselective ring-opening at the methine site, whereas for isobutene oxide/CO₂ copolymerization the regioselective ring-opening occurred at the methylene site (Scheme 10); oxiranes with linear alkyl groups formed copolymers with lower regioselectivity. The glass transition temperatures (T_g) of various oxirane/CO₂ copolymers ranged from −38 to +84 °C (Figure 16) providing the use of similar copolymers as elastomers or plastics.

4.2.4. Copolymerization of Oxiranes with Other C1 Monomers

COS and CS₂ are analogs of CO₂, and the interest towards their using as C1 comonomers is fully justified. However, there are some sporadic publications in this field. The basic feature of copolymerization of oxiranes with COS or CS₂ is O/S exchange side reaction with a formation of copolymers of rather complex composition and mixtures of cyclic thiocarbonates (which was observed in PO copolymerizations, catalyzed by Zn/Co DMCC [161]); under the action of DMCCs, CS₂ and oxirane can transform to COS/CS₂ and thiirane, respectively (Scheme 11a).

Nevertheless, in some cases scientists were able to synthesize well-defined copolymers. For example, cyclohexene oxide was copolymerized with COS in the presence of Zn/Co DMCC with a formation of poly(cyclohexene monothiocarbonate)s (Scheme 11b) [162]. These copolymers contained at least 90% monothiocarbonate linkages. They had T_g = 112 °C, initial decomposition temperature of 214 °C, refractive index of 1.705, and might potentially be used as optical material. When using styrene oxide, copolymerization with COS proceeded with high regioselectivity via nucleophilic attack at methine carbon atom of oxirane with a formation of –(CHPhCH₂SC(O)O)_n– [163]. The results on catalytic copolymerization of oxiranes with COS were summarized and discussed in review of Luo, Zhang, and Darensbourg [164].

In contrast with Zn/Co DMCCs, the well-defined complex Co₃[Co(CN)_6]2 catalyzed copolymerization of PO with COS with a formation of poly(propylene monothiocarbonate)s (M_n = 66.4–139.4 kDa, Ð_M = 2.0–3.9) with suppressed oxygen–sulfur exchange reaction, the selectivity of the monothiocarbonate over carbonate linkages was >99%. The side product, cyclic thiocarbonate, was not formed [165].

4.2.5. Copolymers of Two Different Oxiranes and CO₂

Poly(cyclohexene carbonate) represents brittle plastic, which significantly limits its use. In recent study of Liu et al., transformation of this copolymer into elastomer was achieved via CO₂/cyclohexene oxide/vinyl cyclohexene oxide terpolymerization followed by thiol–ene click reaction and Cu–S coordination post-functionalization. The resulting thermoplastic elastomers displayed good elastic recovery (remaining at 86% strength after five cyclic tests), a certain hardness (Shore A from 31 to 91), and adjustable mechanical properties [166].

4.2.6. Functional Derivatives of PECs

PECs, obtained by copolymerization of PO and CO₂ with different molecular architectures, determined by the used initiators (benzenedi-, tri-, or tetracarboxylic acid), were functionalized by the product of reaction of 2-amino-4-hydroxy-6-methylpyrimidine with OCN(CH₂)₆NCO (Scheme 12); 0.1 mm thick films, prepared by compression molding, demonstrated different mechanical and adhesive properties, determined by copolymer architecture and microstructure [152]. The highest shear strengths s in lap joint tests was demonstrated by copolymer obtained using 1,3,5-benzenetricarboxylic acid as an initiator (F_CO2 = 44%), the values of s were 7.5 MPa (stainless steel) and ~10 MPa (wood).

In light of future use of PEC polyols in polyurethane production, in 2024, Dong et al. proposed the use of primary aromatic amines as an initiators of PO/CO₂ copolymerization, catalyzed by Zn/Co DMCC [167]. The ability of anilines to initiate copolymerization was determined by their pK_b: 2-chloroaniline and 2-bromoaniline (pK_b = 11.35 and 11.47, respectively) were the most efficient initiators. Isomeric nitroanilines with close pK_b were inactive due to coordination of –NO₂ group at catalytic centers. Low-MW amino PEC polyols, obtained using 4,4′-methylene-bis(2-chloroaniline) as an initiator, were suitable for the preparation of rigid polyurethane foams without any external catalyst.

An interesting and topical pathway of post-modification of PEC polyols, prepared using Zn/Co DMC catalysts, was the introducing of a cyclic carbonate end-group (Scheme 13). A copolymer modified in this manner was mixed with cyclic-carbonated soybean oil, and the mixture was then cross-linked by propane-1,3-diamine with a formation of “non-isocyanate” polyurethane [168].

4.3. Other Copolymers Based on Oxiranes

4.3.1. Copolymerization of Oxiranes with Other Monomers

Introduction of reactive fragment into backbone of the macromolecule represent common and efficient approach to new materials; therefore, the combination of polyether fragments (contribution of oxirane) with unsaturated C=C bonds or functional groups from other comonomers appears to be promising, and it seems obvious to use DMC complexes as catalysts for the synthesis of similar copolymers. Back in 2004, PO was copolymerized with maleic anhydride in the presence of Zn/Co DMCC with a formation of alternating copolymer, productivity was up to 10 kg∙g_cat⁻¹, and remarkably, MA was not able to polymerize in the presence of this catalyst [169]. In 2020, similar copolymers attracted the attention of the Zhang’ group: they studied copolymerization of PO and other oxiranes (but-1-ene, hex-1-ene, and cyclohexene oxides, tert-butyl and phenyl glycidyl ethers) with MA, catalyzed by Zn/Co DMCCs [170]. In all cases, copolymers with <1 mol% of the ether content were obtained that had M_n = 47–152 kDa, high values of the temperature of decomposition T_d = 265.7–311.5 °C (5% weight loss), and high stress at break up to 25.3 MPa. Remarkable results were detected when studying copolymers of MA subjected to cis-/trans-isomerization using the previously published method [171]: mechanical characteristics (Figure 17) and biodegradability of the copolymer were significantly improved.

Zn/Cr DMCC (~0.5 wt%) at 110 °C catalyzed copolymerization of PO with phthalic anhydride, and copolymer with M_n = 4.12 kDa and Ð_M = 1.29 was obtained [93], although this material was not studied in terms of mechanical and physical characteristics. But in 2022, Zhang et al. discovered the amazing property of similar linear non-conjugated polyesters: clusteroluminescence [172]. Four nearly alternating copolymers P1–P4 (Figure 18) were prepared using Zn/Co DMCC. The changes of T_g values indicated the decreasing segmental mobility from P1 to P4 at room temperature; copolymer P3 showed the maximum λ_em and the highest quantum yield (QY). The value of QY increased with an increase of the ester content in P3. The occurrence of intensive yellowish-green clusteroluminescence in P3 was explained under the assumption that forbidden n-π* transition of the carbonyl group is allowed in P3 and its analogs.

UV-vis spectral studies confirmed this assumption but revealed longer-wavelength emission peaks at 520–530 nm for P2 and P3, attributed to ester unit clusters, beyond the single C=O group level. From the point of view of application, clusteroluminescence quenching as a result of destruction of polyester clusters under the action of metal ions was illustrated by an example of Fe³⁺ ions: copolymer P3 has shown high selectivity to Fe³⁺, and therefore could be used as an efficient iron sensor.

Further studies [173] of oxirane/anhydride-based polyesters, prepared using Zn/Co DMCCs, revealed clusteroluminescent materials with λ_em ~ 440 nm but with different emission efficiency (Figure 19).

Copolymerization of isobutylene oxide and cyclic anhydrides was studied by Zhang et al. [174]. The screening of different catalysts in copolymerization with MA revealed significant flaws of different homogeneous catalysts manifesting in the formation of low-MW oligomers with high polyether content and side products, isobutyraldehyde and 2-isopropyl-4,4-dimethyl-1,3-dioxolane (Scheme 14a). When Zn/Co DMCC was used, high-MW polyesters with alternating microstructure (>99%) were obtained, and significant amounts of isobutyraldehyde were detected at 80–100 °C. When copolymerization reactions were conducted at 45–60 °C, 45–96% conversions were achieved after 6–9 h for MA, DMMA and PA (see Scheme 14b), copolymerization of SA and F₄PA required 48 and 60 h to achieve 99% conversion. A 99% regioselectivity was observed for SA, PA and F₄PA. Obtained copolymers represented semicrystalline polyesters (T_m = 67–141 °C) depending on the structures of anhydrides. The cis-/trans- isomerization of the C=C bond in the backbone of poly(IBO-alt-MA) was accompanied by a change of T_m from 72 to 153 °C.

The direct (i.e., statistical) copolymerization of cyclic siloxane monomers with oxiranes is impeded because of the insufficient nucleophilicity of the active alkoxide chain end of polyethers for the cleavage of a Si–O bond. Surprisingly, the Zn/Co DMC complex proved to be a capable catalyst for direct copolymerization of PO with (Me₂SiO)₃ at 120 °C [175]. During copolymerization, fast consumption of PO was observed during at the initial stage, whereas ROP of (Me₂SiO)₃ occurred when PO was consumed to nearly full conversion. As a result, gradient copolymers with T_g from −67 °C (3 mol% of Me₂SiO) to −95 °C (46 mol% of Me₂SiO) were formed.

4.3.2. Copolymers of Oxiranes, CO₂ and Anhydrides

The idea of the use of additional comonomers for improvement of mechanical characteristics of PCs and PECs or to create an opportunity for post-functionalization of copolymers has been investigated in numerous studies [6,176,177,178]. Preparation of cross-linkable copolymers via copolymerization of PO, CO₂, maleic anhydride, and allyl glycidyl ether was conducted by Müller et al. in 2016, and allyl-substituted oxirane was introduced initially or at a late stage of copolymerization [179]. Subsequent UV curing with the use of (PhCO₂)₂ as an initiator resulted in transparent films of interest to coating applications.

A three-component catalyst comprising zinc glutarate, rare earth ternary complex, and Zn/Co DMC complex was proposed for the use in terpolymerization of PO, CO₂, and trimellitic anhydride [71]. Under optimized conditions, the yield of copolymer was 66 g∙g_cat⁻¹. In large part, this copolymer had alternating microstructure (Scheme 15) and was characterized by M_w up to 83 kDa, unique thermal stability, and high decomposition temperature (10% loss of weight at 313 °C).

Terpolymerization of PO, CO₂ and itaconic anhydride, catalyzed by Zn/Co DMCC (prepared using ^tBuOH as a CA), was complicated by isomerization of double bond in itaconic anhydride and cross-linking with a formation of gels [180]. The mechanism of cross-linking, presented in this work, seems not quite clear.

4.3.3. Copolymers of Oxiranes, CO₂ and Cyclic Esters

In 2015, the introduction of ε-caprolactone (ε-CL) units to a copolymer, containing ether and carbonate fragments, was proposed with the use of cross-chain exchange reaction (CCER) approach, a combination of two independent chain propagations generated by two different catalysts [181]; in this work, Zn/Co DMCC and tin(II) 2-ethylhexenoate (SnOct₂) were applied for copolymerization of cyclohexene oxide, CO₂, and ε-CL with a formation of materials that had wide spectrum of mechanical properties and optical characteristics.

Copolymers of PO, CO₂, and dl-lactide (dl-LA), prepared in the presence of Zn/Fe DMCCs and synthesized by mechanochemical method in the presence of quaternary ammonium salts, had high dl-LA comonomer incorporation [83]. The further prospects of the use of similar catalysts seem doubtful due to very low catalytic activity (up to 5.65 g∙g_Zn⁻¹∙h⁻¹).

Finally, note that the very idea of copolymerization of three types of comonomers with fundamentally different abilities to form metal–comonomer intermediates and mechanisms of ROP seems overly hopeful, bearing in mind diversity and sensitivity of the ROP coordination catalysts to the structure of the monomers [182].

4.4. Ring-Opening Polymerization of Cyclic Substrates Different from Oxiranes

Less strained (comparative to oxiranes) cycles including cyclic esters (lactones, lactides) were also studied in polymerization, catalyzed by DMCCs. In 2020, Liu et al. showed that Zn/Co nanolamellar DMCC (prepared using ^tBuOH as a CA [160]) in the presence of ~1 mol% of oxiranes initiates ROP of ε-CL in bulk at 120 °C with a formation of high-MW polymer (M_n up to 178 kDa) [94]. The results of this study are of interest considering that the ROP was efficiently initiated by cyclohexene oxide, whereas conventional ROP initiator BnOH was inactive. The following year [63], Kim et al. conducted a comparative study of ε-CL polymerization in the presence of Zn/Co DMCCs (0.33 mol% Zn), prepared with the use of ethyl acetoacetate, ^tBuOH, or Me₂CHCN as CAs; the catalyst prepared with the use of Me₂CHCN showed the highest productivity. Depending on the initiator used (ethylene glycol, propylene glycol, glycerol, or sorbitol), polymers with M_n = 0.75–4.0 kDa were obtained (SEC data), allowing them to be used in the preparation of poly(ester urethane) elastomers. Similar catalytic activities in polymerization of ε-CL and δ-valerolactone were demonstrated by Zn/Co DMCCs, prepared with the use of CH₃NO₂ or 1-methylpyrrolidin-2-one [65]. The effect of the CA on catalytic activity of Zn/Co DMCCs in ε-CL polymerization was also demonstrated in the comparative study of the complexes, prepared using different aliphatic nitriles that confirmed the highest efficiency of Me₂CHCN [64].

Besides polyurethane production, low-MW ε-CL-based polyols (M_n = 1.0–4.2 kDa, Ð_M = 1.30–1.88), prepared with the use of Zn/Co DMCCs, were employed as soft segments to produce thermoplastic poly(ester–ester) elastomers containing poly(butylene terephthalate) as a hard segment [183].

4.5. Other Copolymers

The potential ability to fixation of CO₂ by its interaction with oxiranes with a formation of cyclic carbonates (for more details see Section 4.6) allowed Liu et al. [184] to develop an original one-pot synthetic approach to poly (2-oxo-1,3-dioxolane-4-yl)methyl methacrylate, based on glycidyl methacrylate (Scheme 16). This approach was based on simultaneous use of a Zn/Co DMCC and [C₁₆H₃₃NMe₃]Br catalytic system (0.3 mol% related to oxirane) and a free-radical initiator in DMF media at appropriate temperature. Up to 100 mol% of the carbonate content in polymers were achieved. UV-vis spectra of these copolymers showed high absorbance at 200–315 nm (UB-B and UV-C regions) and 93% transmittance (0.04 mm thick) for UV-A and visible regions, which makes these copolymers prospective UV-absorbing material for healthcare applications.

4.6. Synthesis and Transesterification of Cyclic Carbonates

4.6.1. Synthesis of Cyclic Carbonates

Interaction of oxiranes with CO₂ may lead to the formation of five-membered cyclic carbonates, this catalytic process attracts researchers’ attention as one of the possible ways of chemical fixation of CO₂ with a formation of value-added products [185,186,187]. Zn/Co DMCCs, being efficient catalysts of copolymerization of oxiranes with CO₂, can also be tuned to the formation of cyclic carbonates. In 2009, Park et al. reported the results of the study of addition of quaternary ammonium salt to Zn/Co DMCCs [188]. They showed that [R₄N]X (R = ⁿPr, ⁿBu, ⁿHex, ⁿOct, n-C₁₂H₂₅; X = Cl or R = ⁿBu, X = Br) promote the formation of cyclic carbonates, 140 °C was the best reaction temperature condition at moderate pressure of CO₂ (3.5 bar), and TON values up to 10³ and selectivity of 99% were achieved. In the presence of Zn/Co DMCC (3 wt%) and [ⁿBu₄N]Br (1 wt%), 96% conversion of PO was achieved after 20 h at 60 °C and 20 bar of CO₂ [189]. A continuation of Park’s group research [69] was an attempt to overcome the disadvantages of using two different catalysts (Zn/Co DMCCs and [R₄N]X) by introducing of [C₁₆H₃₃NMe₃]Br as a CA during the preparation of the catalyst. Under optimized conditions, a number of oxiranes were converted to corresponding cyclic carbonates (

Table 4. Cycloaddition of CO₂ and various oxiranes, catalyzed by Zn/Co DMC complex, prepared using [C₁₆H₃₃NMe₃]Br as a complexing agent [69] ¹.

Entry	Oxirane	Cyclic Carbonate	Time, h	Conv., %	Yield, %	TON 2
1			6	82	82	3075
2			6	93	92	3486
3			6	95	91	3563
4			6	86	83	3225
5			4	92	91	3450
6			12	47	46	1763

¹ Reaction conditions: oxirane = 18.6 mmol, DMCC amount = 20 mg, p(CO₂) = 12 bar. ² Turnover number (TON): mol of oxirane converted per mol of Zn.

). The reaction mechanism, proposed in [69], does not hold water. The same catalytic system was successfully used for the synthesis of bis-carbonates from bis-epoxides containing two glycidyl fragments with different spacers [190].

When studying copolymerization of PO with CO₂ using Zn/Co DMCCs, Zhang et al. observed that the trace amounts of water facilitate the formation of cyclic carbonates [191]. In their further study, they showed that binary catalytic systems, containing conventional Zn/Co DMCCs prepared using ^tBuOH and P123, in combination with quaternary ammonium salts, in the presence of water (260 ppm) provide formation of the corresponding cyclic carbonate with ~100% selectivity; the maximum TON was 6300 [51].

Zn/Fe DMCCs, prepared by ball milling at 140 °C and 30 bar of CO₂ after 6 h of the reaction between PO and CO₂, showed TON up to 2300, and the highest yield of propylene carbonate was 97.3% [80,81].

In 2022, the study of DMC M₃[Co(CN)₆]₂ (M = Fe, Co) in combination with [ⁿBu₄N]Br showed that Co/Co complex is a selective catalyst of the formation of cyclic carbonates [192]; the yields (1 mg of the catalyst, 25 mmol of oxirane, 65 °C, 10 bar, 7 h) exceeded 99% for PO, 1,2-butylene oxide and epichlorohydrin. The yields of carbonates significantly decreased with increasing of steric hindrance, as indicated by formation of 3-tert-butoxy-1,2-propylene carbonate, 1-phenyl-1,2-ethylene carbonate, and 3-phenoxy-1,2-propylene carbonate with yields of 93, 64, and 66%, respectively. The suggested mechanism of the reaction involves coordination of the oxirane oxygen atom at the Co Lewis acid site, nucleophilic attack of Br⁻ on the less sterically hindered C atom, coordination/insertion of CO₂, and intramolecular cyclization (Scheme 17).

The same year, Znang et al. demonstrated that even Prussian blue Fe₄[Fe(CN)₆]₃ mesoporous nanoparticles, surface-functionalized by –OH groups, provide quantitative yields of cyclic carbonates under mild reaction conditions (90 °C, 10 bar of CO₂) [193].

Glycerol carbonate (4-hydroxymethyl-2-oxo-1,3-dioxolane) is one of the most valuable bifunctional building blocks in organic chemistry and polymer science [194,195]. Conversion of glycidol to glycerol carbonate is a difficult task, complicated by the formation of oligomeric and polymeric products. Comparative study of DMC complexes, prepared without the use of CAs, revealed the most active system, Zn₃[Co(CN)₆]₂∙5H₂O [66]. When using 0.02 mol% of the catalyst at 120 °C and 20 bar, 97% conversion of glycidol and 71% yield of glycerol carbonate were achieved.

4.6.2. Transesterification of Cyclic Carbonates

Methanolysis of cyclic carbonates to produce dimethyl carbonate (Scheme 18a) is a green and atom-economical alternative to conventional method of the synthesis of (MeO)₂C=O from phosgene; this approach can also compete with the Enichem catalytic process (Scheme 18b) [196,197].

In 2006, the use of Zn/Fe DMCC was proposed for methanolysis of propylene carbonate, isolated yield of (MeO)₂C=O was 86% [198]. Further comparative study of different DMCCs (Zn, Mn, Ni/Fe and Zn, Mn, Ni/Co) [199] revealed the most efficient catalyst for this process, a Mn/Fe DMC complex, prepared by the reaction of K₄Fe(CN)₆ with MnCl₂ at 1:8 molar ratio. At 140 °C and catalyst loading 5 g per mol of propylene carbonate, the TOF was 57.1 h⁻¹ and selectivity of (MeO)₂C=O formation was 96.1%. The loss of catalytic activity during the recycling process was 11.2% after four runs. The rate of ethylene carbonate conversion in the presence of Mn/Fe DMC was higher compared to propylene carbonate. For this catalyst and process, a simple two-step power low kinetic model was developed; good correlations with experimental data were observed [121].

4.7. The Use of DMCCs in Preparative Organic Chemistry

4.7.1. Hydroamination

Hydroamination of terminal alkynes is of interest to organic chemistry, and DMCCs were studied in this reaction. For example, reaction of 4-isopropylaniline with PhC≡CH was catalyzed by Zn/Co DMCC prepared using poly(tetramethylene ether) glycol as a CA, and 99% yield of the hydroamination product was achieved after 24 h at 110 °C with 100 mg_cat∙mmol⁻¹ loading [41]. In a silica-supported Zn/Co DMC catalyzed reaction of 4-isopropylaniline with PhC≡CH in toluene at 110 °C, the TOF was only ~0.6 h⁻¹ [73]. A well-defined DMC complex Zn₂[Co(CN)₆](OAc)∙4H₂O showed higher catalytic activity in hydroamination of PhC≡CH by 2,6-diisopropylaniline (TOF = 0.87 h⁻¹) in comparison with conventional Zn/Co DMCCs [60]. A series of Zn/Co DMCCs, prepared with the use of different CAs and co-CAs, also showed moderate activities; aliphatic amines have proven to be less reactive [40].

Recently performed experiments on Zn/Co DMC-catalyzed hydroamination of substituted phenylacetylenes by substituted anilines confirmed low productivity of these catalysts; at 5 wt% loadings, the yields of the corresponding imines were up to 94% after 12 h at 110 °C [200]. The same yields were achieved in the reaction of PhNHNH₂ with PhC≡CH.

4.7.2. Reductive Amination

Zn/Co DMCC, prepared using ^tBuOH and PEG-10000 and pre-activated at 180 °C, demonstrated moderate activity in reductive amination of ketones and aldehydes by substituted anilines; BnNH₂ (10 wt% of the catalyst, 60 °C, 5 h, MeOH as a solvent), poly(methylhydrosiloxane) was used as a reducing agent, and the yields of amines amounted to 52–98% [201]

4.7.3. Transesterification

Fatty acid methyl esters (FAMEs) are intermediates for the production of fine oleochemicals; mostly FAMEs are considered as a fuel (biodiesel) [202]. Industrial production of FAMEs via transesterification of natural oils and fats is based on the use of homogeneous catalysts; however, heterogeneous systems keep attracting the attention of researchers due to the ease of separation and recyclability [203,204,205,206]. In 2020, Zn/Fe DMCC was studied in methanolysis of jatropha oil; 95% conversion was achieved at [MeOH]/[Oil] = 10:1, 160 °C and 5 wt% catalyst loading [207].

4.7.4. Acid-Nitrile Exchange

The synthesis of nitriles from carboxylic acids can be performed by a thermodynamically-driven reaction between acid and glutaronitrile (Scheme 19). This thermally induced process proceeds without any additional catalyst with ~50% yields (2-phenylbutyric acid as a substrate, 215 °C, 16 h), but some inorganic materials including DMC complexes can accelerate the reaction. The best results among other DMCCs was demonstrated by Co(II)–Co(III) system (86%), and homogeneous catalyst AlCl₃ turned out to be equally active (90% yield); the catalyst loadings were 10 mol% [208].

4.7.5. Synthesis of Phosphoramidates

Cu₃[Co(CN)₆]₂-based DMCC in the presence of I₂ as an additive proved capable of catalyzing oxidative transformation of phosphites to phosphoramidates [209]. After optimization of the reaction conditions for (ⁿBuO)₂P(O)H/PhCH₂CH₂NH₂ substrates (3 wt% of the catalyst, 30 min, 20 °C, 1:2:0.15 phosphite/amine/I₂ molar ratio in CH₂Cl₂ under O₂ atmosphere), the method was applied to different phosphites and amines (Scheme 20). Catalytic amounts of I₂ provided in situ formation of iodophosphate intermediate, followed by the formation of the P-N bond; the catalytic cycle was closed by aerobic oxidation involving the active center of DMC complex.

4.7.6. Epoxidation

The complex Cu₃[Co(CN)₆]₂ (0.6 mg∙mmol⁻¹) was studied as a catalyst of epoxidation of styrene, cyclohexene, and norbornene using ^tBuOOH as an oxidant; high selectivity of the formation of oxirane was achieved only for norbornene. In the case of styrene up to 64% of PhCHO were formed, depending on the reaction solvent [88].

Similar results in epoxidation of styrene by ^tBuOOH were obtained for wide spectrum of DMC complexes (Co(II)–Fe(III), Co(II)–Fe(II), Co(II)–Co(III), Cu(II)–Fe(III), Ni(II)–Fe(III), Fe(II)–Co(III), Fe(II)–Fe(III), and Fe(II)–Fe(II)), with and without modification by ionic liquids [210]. The complex Co₃[Fe(CN)₆]₂ (6 mg∙mmol⁻¹), modified with 1-butyl-3-methylimidazolium bis(trifluoromethylsulfonyl)imide showed the highest catalytic efficiency (90% conversion after 6 h at 72 °C), but the reaction products contained 35 mol% of PhCHO.

Modification of the structure of Fe₃[Co(CN)₆]₂ by introducing of ROH (R = C1–C6 alkyl) failed to reach any breakthrough in epoxydation of styrene; a slight increase of activity was accompanied by decrease of selectivity of the formation of oxirane [211].

4.7.7. Catalytic Rearrangements of Biomass-Derived Furans

Hydrogenative ring-rearrangement of biomass-derived furfural and 5-hydroxymethyl furfural to other value-added products was the subject of the study of Deng et al. [78]. They showed that Pd-Zn/Fe DMCCs demonstrate high efficiency for the synthesis of cyclopentanone or 3-hydroxymethyl cyclopentanone with yields up to 96.6 and 87.5%, respectively, whereas the Pd-Ni/Fe and Pd-Co/Fe DMCCs with a weak Lewis acidity provide ~90% yields of furfuryl alcohol or 2,5-bis(hydroxymethyl)furan, respectively. The catalyst loadings were ~10 mg∙mmol⁻¹ (150 °C, 40 bar of H₂, aqueous media), Pd-Zn/Fe DMCCs demonstrated excellent reusability after simple washing by water. In the absence of H₂, Zn/Fe DMC complex efficiently catalyzed transformation of 5-hydroxymethyl furfural to corresponding cyclopentenone (Scheme 21); it is noteworthy that in the presence of H₂, furfural transformed to cyclopent-2-enone [212]. The authors proposed that Zn atoms represent acid and hydrogenation active sites, i.e., Zn/Fe DMCC is a rare example of Zn-based hydrogenation catalyst.

Conversion of furane derivatives to cyclopentanones and cyclopentenones represents a promising “green” alternative to conventional decarboxylative ketonization of adipic acid. Comparison of different catalytic systems, studied in conversion of furfural and 5-hydroxymethyl furfural to cyclopentanones [213], revealed the undoubted advantage of Pd-Zn/Fe DMCCs, which were inferior in selectivity only to Au/TiO₂ catalyst. In this way, DMCCs can be used to produce highly valuable furan-based chemicals from biomass, and this direction is promising and worthy of further development.

5. Conclusions and Prospects

The current use and prospective applications of DMCCs are presented in generalizing Scheme 22, which also reflects the type of the raw material, fossil or renewable. Among other DMCCs, Zn/Co DMC complexes remain the most active and widely used catalysts for the production of relatively low-MW polyols, products of ROP of propylene oxide (Scheme 22a), and other oxiranes to a lesser extent. Chemical fixation of CO₂ in polyether backbone meet the requirements of contemporary green agenda and allows obtaining poly(ethers-co-carbonate)s (Scheme 22b), polyols no less in demand. Both copolymers are widely used in polyurethane industry. A deeper understanding of the mechanism of the action of DMCCs allows developing active catalysts with new formulations, morphologies, and properties; however, other DMC complexes distinct from Zn/Co DMCCs are not yet ready to compete with the latter in these industrially implemented processes. The mechanism of the catalytic action of DMC complexes in the processes, presented in Scheme 22a,b, is still not clear. Theoretical and fundamental studies are based on idealized models, and both mononuclear and binuclear concepts are still being discussed. Hopefully, the current progress in quantum chemical modeling and physico-chemical monitoring of the catalytic processes will clarify the mechanisms of the action of DMC complexes in the near future.

As stated above, an alternative method of chemical fixation of CO₂ is a transformation of oxiranes to cyclic carbonates (PCs), highly demanded solvents and reagents (Scheme 22c). In this reaction, DMC complexes of other metals, distinct from conventional Zn/Co DMCCs, can be used. At the same time, development of new types of Zn/Co DMCCs with new complexing agents or via deeper design of the catalyst’s morphology (layered materials) also open up new growth avenues.

Besides preparation of polyols, PECs and PCs, DMCCs are increasingly used in the synthesis of new types of polymeric materials (Scheme 22d) promising for their applications as adhesives, elastomers, sensors, and UV-absorbing materials. Currently, there is a growing interest in the use of DMCCs in fine organic chemistry, and their applications for (hydro)amination, transesterification, and catalytic processing of renewable raw materials such as biobased furan derivatives (Scheme 22e) and triglycerides (Scheme 22f). As can be seen from data presented in Section 4.7, the catalytic potential of DMCCs in synthetic organic chemistry and biomass valorization is still not fully disclosed. These green, sustainable, and efficient processes need be further explored in the future.