**Highly E**ffi**cient and Reusable Alkyne Hydrosilylation Catalysts Based on Rhodium Complexes Ligated by Imidazolium-Substituted Phosphine**

#### **Olga Bartlewicz 1,\*, Magdalena Jankowska-Wajda <sup>1</sup> and Hieronim Maciejewski 1,2**


Received: 15 April 2020; Accepted: 28 May 2020; Published: 1 June 2020

**Abstract:** Rhodium complexes ligated by imidazolium-substituted phosphine were used as catalysts in the hydrosilylation of alkynes (1-heptyne, 1-octyne, and phenylacetylene) with 1,1,1,3,5,5, 5-heptamethyltrisiloxane (HMTS) or triethylsilane (TES). In all cases, the above complexes showed higher activity and selectivity compared to their precursors ([Rh(PPh3)3Cl] and [{Rh(μ-Cl)(cod)}2]). In the reactions with aliphatic alkynes (both when HMTS and TES were used as hydrosilylating agents), β(Z) isomer was mainly formed, but, in the reaction of phenylacetylene with TES, the β(E) product was formed. The catalysts are very durable, stable in air and first and foremost insoluble in the reactants which facilitated their isolation and permitted their multiple use in subsequent catalytic runs. They make a very good alternative to the commonly used homogeneous catalysts.

**Keywords:** hydrosilylation; alkynes; heterogeneous catalysis; rhodium catalysts; ionic liquids

#### **1. Introduction**

Vinylsilanes and siloxanes due to a relatively low cost of their synthesis, low toxicity, and good chemical stability are valuable raw materials applied in many organic syntheses such a nucleophilic substitution, alkylation [1,2], and coupling [3,4]. However, the course of the above syntheses and formation of desired products are influenced by the kind and purity of isomer of vinyl organosilicon derivative employed. This is why various methods are developed for the synthesis of the aforementioned derivative to ensure high regio- and stereoselectivity of vinylsilanes and siloxanes formed. One of the most popular and commonly applied (also on a commercial scale) methods of synthesis of organofunctional silicon compounds is hydrosilylation [5,6], which enables obtaining vinyl derivatives in the reaction with alkynes. However, the reaction course and the kind of products formed depend on many factors such as the type of alkyne (terminal or internal) and hydrogen silane (siloxane) as well as reaction conditions (temperature, time, the kind of solvent) and particularly the kind of catalyst used [7]. In the case of hydrosilylation of terminal alkynes, there is a possibility of the formation of three isomers β-(E), β-(Z), and α as shown in Scheme 1:

$$\stackrel{\mathsf{R}}{=} \stackrel{\mathsf{\mathsf{i}}}{} \stackrel{\mathsf{\mathsf{i}}}{} \stackrel{\mathsf{\mathsf{i}}}{} \stackrel{\mathsf{\mathsf{i}}}{} \stackrel{\mathsf{\mathsf{i}}}{} \stackrel{\mathsf{\mathsf{i}}}{} \stackrel{\mathsf{\mathsf{i}}}{} \stackrel{\mathsf{\mathsf{i}}}{} \stackrel{\mathsf{\mathsf{i}}}{} \stackrel{\mathsf{\mathsf{i}}}{} \stackrel{\mathsf{\mathsf{i}}}{} \stackrel{\mathsf{\mathsf{i}}}{} \stackrel{\mathsf{\mathsf{i}}}{} \stackrel{\mathsf{\mathsf{i}}}{} \stackrel{\mathsf{\mathsf{i}}}{} \stackrel{\mathsf{\mathsf{i}}}{} \stackrel{\mathsf{\mathsf{i}}}{} \stackrel{\mathsf{\mathsf{i}}}{} \stackrel{\mathsf{\mathsf{i}}}{} \stackrel{\mathsf{\mathsf{i}}}{} \stackrel{\mathsf{\mathsf{i}}}{} \stackrel{\mathsf{\mathsf{i}}}{} \stackrel{\mathsf{\mathsf{i}}}{} \stackrel{\mathsf{\mathsf{i}}}{} \stackrel{\mathsf{\mathsf{i}}}{} \stackrel{\mathsf{\mathsf{i}}}{} \stackrel{\mathsf{\mathsf{i}}}{} \stackrel{\mathsf{\mathsf{i}}}{} \stackrel{\mathsf{\mathsf{i}}}{} \stackrel{\mathsf{\mathsf{i}}}{} \stackrel{\mathsf{\mathsf{i}}}{} \stackrel{\mathsf{\mathsf{i}}}{} \stackrel{\mathsf{\mathsf{i}}}{} \stackrel{\mathsf{\mathsf{i}}}{} \stackrel{\mathsf{\mathsf{i}}}{} \stackrel{\mathsf{\mathsf{i}}}{} \stackrel{\mathsf{\mathsf{i}}}{} \stackrel{\mathsf{\mathsf{i}}}{} \stackrel{\mathsf{\mathsf{i}}}{} \stackrel{\mathsf{\mathsf{i}}}{} \stackrel{\mathsf{\mathsf{i}}}{} \stackrel{\mathsf{\mathsf{i}}}{} \stackrel{\mathsf{\mathsf{i}}}{} \stackrel{\mathsf{\mathsf{i}}}{} \stackrel{\mathsf{\mathsf{i}}}{} \stackrel{\mathsf{\mathsf{i}}}{} \stackrel{\mathsf{\mathsf{$$

The kind of isomer formed depends on the way of ≡Si–H addition to the C≡C triple bond. In the former two cases, the silicon atom bonds to the terminal carbon atom which results in β-(E) isomer (if *cis*-addition occurs) or β-(Z) isomer (if *trans*-addition occurs). On the other hand, the reverse addition leads to the geminal vinyl silane (α isomer) as an internal adduct.

As mentioned above, the catalyst has a significant influence on the reaction course. The catalysts used for alkyne hydrosilylation are diverse simple compounds and complexes of transition metals such as Rh [8], Pt [9,10], Ru [11], Ni [12], Co [13], and Ir [13]. Their activity in many cases is high, but they differ in selectivity [7].

For instance, platinum catalysts (including Karstedt's catalyst), which are among the most popular catalysts for hydrosilylation, in the case of the reaction with terminal alkynes preferentially form β(E) isomer as the main product and, to a small extent, also α isomer. On the other hand, β(Z) isomer is not formed (or formed only in negligible amounts) in the reactions catalyzed by platinum catalysts [14], whereas rhodium catalysts favor mainly the formation of β(Z) product [8], except for cationic rhodium complexes, e.g., [{Rh(cod)2}BF4/PPh3], in the presence of which the hydrosilylation reactions result in the formation of predominant amounts of the β(E) product [15,16]. In the literature, there are reports on the effect of different factors on the stereoselectivity of the reaction of hydrosilylation conducted in the presence of rhodium complexes. They include the kind of the structure of starting compounds [17], the type of solvent [16,18] as well as the way and sequence of adding reactants [19–21]. On the basis of the above data, some conclusions can be drawn that will enable directing the reaction to the desired isomer. For example, the reactions of silanes with electron-donor substituents (e.g. trialkylsilanes) and terminal alkynes not containing bulky substituents (with a large steric hindrance) result mainly in the formation of β(Z) isomer, whereas the reactions of silanes with electron-withdrawing substituents (e.g. alkoxy- or chlorosilanes) lead preferentially to the formation of β(E) isomer [20,22]. This effect is enhanced when an alkyne contains large substituents. Quite a large group of catalysts used in the hydrosilylation of terminal alkynes consists of various rhodium(I)-NHC complexes which in most cases catalyze the selective formation of β(Z) product [23–25]. However, if the steric hindrance caused by substituents both in NHC ligand and alkyne is large, then the formation of mostly β(E) product is possible [26]. The use of complexes with NHC ligands, as well as the addition of other ligands (e.g., phosphines) that are sensitive to various contaminants and are unstable in the presence of oxygen and moisture, results in the necessity of performing the reaction in a closed system in the atmosphere of inert gases while maintaining an appropriate level of purity of the reagents [12,23,27]. This is why researchers keep searching for new catalysts that are characterized by high activity and selectivity as well as stability and resistance to contaminants. An alternative solution can be heterogeneous catalysts.

Our research group has been involved for several decades in the development of new catalysts for hydrosilylation. The application of ionic liquids as immobilizing agents for transition metal complexes is an interesting aspect of the above research [28–34]. In all the cases, ionic liquids played the role of a solvent and immobilizing agent for metal complexes which provided a biphasic system with reagents. All the above systems were very effective in the hydrosilylation of olefins. However, ionic liquids can also be employed in another way, i.a. as structural components of the complexing ions [35,36]. Very recently, we have obtained rhodium and platinum complexes with phosphine ligands that were functionalized with imidazolium ionic liquids [37]. As precursors of the above complexes, we have employed Wilkinson's catalyst or *bis*[chloro(1,5-cyclooctadiene)rhodium(I)]. Catalytic studies of olefin hydrosilylation proved very high activity, durability, and stability of the obtained catalysts as well as the possibility of their multiple use [35].

In this paper, we present the results of the study on the catalytic activity of rhodium complexes with phosphine ligands functionalized with ionic liquids for reactions of hydrosilylation of terminal alkynes. Our study was aimed at determining the effect of the kind of catalyst, olefin, and hydrosilylating agent on the yield and selectivity of the hydrosilylation process as well on multiple use of the same portion of the catalyst.

*Catalysts* **2020**, *10*, 608

#### **2. Results and Discussion**

The reaction of alkyne hydrosilylation was studied according to Scheme 2, using three kinds of alkynes.

**Scheme 2.** The reactions of the hydrosilylation of alkynes studied.

We have employed aliphatic alkynes differing in their chain length, 1-heptyne and 1-octyne, as well as an alkyne that contains an aromatic ring in its structure, namely phenylacetylene. Hydrosilylating agents were 1,1,1,3,5,5,5-heptamethyltrisiloxane (HMTS) and triethylsilane (TES). The former of the compounds, due to its high stability, is frequently employed in the synthesis of organosilicon compounds. Moreover, it is a good model compound for obtaining vinylpolysiloxanes, whereas triethylsilane, due to its electron donor properties is a popular reducing agent and a starting material for obtaining vinylsilanes. As catalysts for the above reactions, we have employed Rh(I) complexes that have been recently obtained by our research group. They contain phosphine ligands in which imidazolium ionic liquid is a substituent. Synthesis of the ligands was described in our earlier paper [37], whereas syntheses of rhodium complexes (**1** and **2**) ligated by these ligands are presented in Scheme 3.

**Scheme 3.** Syntheses of rhodium complexes ligated by imidazolium-substituted phosphine.

For the sake of comparison, the catalytic study also included both precursors of rhodium catalysts, i.e., Wilkinson's catalyst [Rh(PPh3)3Cl] and cyclooctadiene rhodium complex [{Rh(cod)(μ-Cl)}2]. The complexes were highly active for hydrosilylation of alkenes and, for this reason, we tested their effectiveness in the reactions with alkynes. Phosphine complexes (**1** and **2**), contrary to the starting complexes (precursors), are insoluble in reactants. Moreover, they are characterized by high thermal stability [38] and resistance to oxidation and moisture. This is why they can be used without the necessity of reagent purification and under normal conditions (in the presence of air in open systems).

At the initial stage, when optimizing hydrosilylation reaction conditions, we performed test reactions in the presence of Wilkinson's catalyst and determined that the optimal stoichiometric ratio of reactants is [HSi≡]:[RC≡CH] = 1.3:1. The excess of HSi≡ enabled obtaining the considerably higher conversion of alkyne. Moreover, test reactions were conducted at different temperatures and the optimum reaction temperature appeared to be 90 ◦C. For example, when the reaction of 1-octyne with HMTS was carried out at 60 ◦C for four hours, the alkyne conversion was 11%, whereas at 90 ◦C the conversion reached 100% already after one hour. Such results were obtained at the catalyst concentration of 1 <sup>×</sup> 10−<sup>3</sup> mol/L mol HSi≡. For this reason, to compare the activity of all the catalysts studied, the ratio [HSi≡]:[RC≡CH]:[cat] <sup>=</sup> 1.3:1:2 <sup>×</sup> <sup>10</sup>−<sup>3</sup> was used. Results of our earlier studies of the reaction of hydrosilylation conducted in the presence of commercially available rhodium catalysts have shown that the optimal reaction time is one hour. However, to verify these results in the presence of the catalysts tested in the present study, we carried out FT-IR in situ analysis for two catalytic systems (the complex **1** and the Wilkinson's catalyst) that enabled us to monitor the reaction course in real time. As test reactions, we have chosen reactions of HTMS and 1-octyne. During the experiments, we kept track of the decay of the HSi<sup>≡</sup> band originated from HMTS (913 cm−1) and the increase in the band originated from the C=C bond (1600 cm<sup>−</sup>1) that shows the formation of the hydrosilylation reaction product (Figure S41 and Figure S42 in Supplementary Materials). To determine the reaction profiles, we have used the HSi≡ conversion calculated as a change in the area of the HSi≡ band (Figure 1). Non-stoichiometric amount of the above reactant, which is also shown by the presence of the remainder of the HSi≡ band (Figure S41 and Figure S42), was taken into account during the determination of the reaction profiles and converted to the stoichiometric amount.

**Figure 1.** The dependence of the conversion of Si–H on time for the reaction of 1-octyne with 1,1,1,3,5,5,5-heptamethyltrisiloxane in the presence of catalyst **1** and Wilkinson's catalyst.

The measurements showed that the new complex **1** makes it possible to reach full conversion considerably faster than the Wilkinson's catalyst. In the case of the complex 1, full conversion was reached as early as after 15 min, whereas, in that of Wilkinson's catalyst, it took 55 min. Therefore, to compare the activity of all complexes in the same conditions and to achieve the highest conversion, we decided to conduct all the further reactions at 90 ◦**C** for 1 h in air in the open system without a solvent. The reactions of alkynes with HMTS were performed first. After the reaction completion, the catalyst was in the form of a fine suspension; therefore, its separation from the post-reaction mixture was carried out by centrifugation. After the separation, the post-reaction mixture was analyzed chromatographically and the obtained results are presented in Table 1.

Chromatographic and spectroscopic analyses of the post-reaction mixture showed the presence of isomers β*(Z)*, β*(E)* and α. The obtained results confirmed the high catalytic activity of rhodium phosphine complexes (**1**, **2**) in all the reaction systems (Table 1). The conversion in the reaction of hydrosilylation of aliphatic alkynes has been considerably higher (100%) than in the reaction with phenylacetylene (87%). Some surprise is relatively low activity and selectivity (and its lack in the case of the reaction with phenylacetylene) of the catalyst [{Rh(μ-Cl)(cod)}2] since this complex belongs to the most active catalysts for hydrosilylation of alkenes [6,7]. New rhodium complexes (**1** and **2**) do not lead to the formation of α isomer, but only to β isomers. For the majority of the complexes studied (except for the catalyst [{Rh(μ-Cl)(cod)}2]), *cis* isomer is overwhelmingly formed, which is

in agreement with results obtained for other rhodium complexes. However, selectivities obtained in reactions conducted in the presence of the catalysts **1** and **2** are higher than those observed in the case of reactions catalyzed by precursors of the rhodium catalysts. One can also notice that, in the case of aliphatic alkynes, the selectivity towards β(Z) isomer increases with the chain growth. It is modest growth, albeit the chain lengths of both alkynes differ only by one carbon atom. When comparing selectivities of the catalysts **1** and **2**, one can observe that in each case the catalyst **1** is characterized by a higher selectivity. To a large extent, it can be explained by differences in the structure and steric hindrance of both complexes. Taking into consideration the mechanism of catalysis occurring in analogous phosphine and cyclooctadiene complexes [7], one can assume that, in the case of the complex **1**, the activation of the catalyst occurs through the detachment of one of the PPh3 ligands, whereas in that of the complex **2** the rupture of one coordinative bond in the cyclooctadiene group occurs. Thereby, the steric hindrance in the complex **2** is greater than that in the complex **1** which results in a decrease in the selectivity and the formation of a large amount of the *trans* isomer.


**Table 1.** Conversion of alkynes and selectivity for isomers in the hydrosilylation of RC≡CH with HMTS.

[RC≡CH]:[HSi≡]:[cat] = 1:1.3:2 × 10<sup>−</sup>3; T = 90 ◦C, t = 1 h.

The catalytic study on the reaction between phenylacetylene and HMTS has shown that the catalyst **1** (where Wilkinson's catalyst was the metal precursor) was characterized by a higher selectivity towards the *cis* product (100%) compared to that of its precursor (in the case of the precursor the *trans* product was formed as well). Hydrosilylation of phenylacetylene catalyzed by the complex **2** also results in the formation of the *cis* product; however, in that case, the selectivity was 86%. The formation of predominant amounts of the *cis* isomer is frequently observed for hydrosilylation reactions catalyzed by neutral rhodium complexes [6,7].

Analogous measurements of catalytic activity were carried out for hydrosilylation of alkynes with triethylsilane. In addition, in this case, the catalysts **1** and **2**, contrary to their precursors, were insoluble in the reactants; therefore, after the reaction, the catalyst was isolated from the mixture by centrifugation. Then, the composition of the post-reaction mixture was determined by using chromatographic analysis and results of the determination are presented in Table 2.


**Table 2.** The conversion of alkynes and selectivity to isomers in the hydrosilylation of RC≡CH with TES.

[RC≡CH]:[HSi≡]:[cat] = 1:1.3:2 × 10<sup>−</sup>3; T = 90 ◦C, t = 1 h.

Based on the performed measurements, one can say that in all the cases the reactions catalyzed by the complexes **1** and **2** resulted in higher yields than those catalyzed by the precursors. In the case of hydrosilylation of aliphatic alkynes, an analogous tendency is observed as that occurring for the

reactions with HMTS, i.e., that the selectivity towards the formation of *cis* isomers increases with the growth of chain lengths. In addition, a bit higher effectiveness of the complex **1**, compared to the complex **2**, was established. A relatively low activity was also confirmed and, in particular, low selectivity in the reactions catalyzed by the complex [{Rh(μ-Cl)(cod)}2]. However, the most interesting results were observed in the case of the reaction between TES and phenylacetylene. Apart from the total lack of activity of the complex [{Rh(μ-Cl)(cod)}2], in other cases, a change in the selectivity to the predominant formation of the *trans* isomer was found. As it was mentioned earlier, the preference for the formation of a particular isomer is determined by steric and induction effects. On the one hand, the phenyl group makes some steric hindrance, but in the reaction with HMTS, which is also a bulky molecule, the *cis* isomer is preferentially formed. In this case, the steric effects are enhanced by the induction effects because triethylsilane is a strong electron-donor agent and, as a result of the interaction with phenyl group (from phenylacetylene), it favors the formation of the *trans* product. Similar effects were observed in the reaction of hydrosilylation of phenylacetylene with dimethylphenylsilane in the presence of Rh-NHC complexes [20]. The predominant formation of the *trans* product is not surprising in the case of platinum catalysts, but in that of rhodium catalysts, it is a rare case [8,16].

As was already mentioned, rhodium complexes (**1** and **2**) were insoluble in the reactants what enabled their isolation from the post-reaction mixture. However, due to the high dispersion of the catalyst which formed a fine suspension and sedimented very slowly, the separation of the catalyst from the post-reaction mixture was ineffective. For this reason, we employed a rotary centrifuge which enabled the fast separation of the catalyst from the mixture. The isolated catalyst was reused in the subsequent catalytic runs, and its activity was determined. The isolation and reusing were performed for the complex **1** which was characterized by higher activity and selectivity in all the above reactions. As a test reaction, we chose hydrosilylation of 1-octyne with HMTS. After each catalytic run, the mixture was centrifuged, followed by the collection of products (colorless and clear liquid) with a syringe. The products were subjected to chromatographic and spectral analyses and the isolated catalyst was reused in a next catalytic run carried out after a fresh portion of the reaction substrates was added. In this way, we conducted five catalytic runs and the obtained results are presented in Table 3.


**Table 3.** The conversion and selectivity for hydrosilylation of 1-octyne with HMTS in subsequent catalytic runs catalyzed by the same portion of the catalyst **1** isolated from the post-reaction mixture by centrifugation.

[RC≡CH]:[HSi≡]:[cat] = 1:1.3:2 × 10<sup>−</sup>3; T = 90 ◦C, t = 1 h.

The results show that catalytic activity of the complex **1** in subsequent catalytic runs is very high, particularly in the first three runs. In the further runs, the conversion of 1-octyne decreases which can be explained by a gradual loss of catalyst as a result of incomplete centrifugation, i.e., only partial catalyst isolation. The slow sedimentation of the catalyst was the reason for which centrifugation was carried out only for a few minutes. The evaluation of the centrifugation effectiveness was based on the visual observation of the separation of two phases. This appeared to be not quite appropriate. This is why we decided to check it by using a different means of isolation. The post-reaction mixture was subjected to filtration followed by placing the filter with the precipitate in the reaction mixture and repeating the catalytic run. The conversion and selectivity in the subsequent runs are presented in Table 4. No significant reduction in the conversion was observed, contrary to the case when the catalyst was reused after centrifugation. This allows supposing that the decrease in the conversion observed previously was the effect of the incomplete isolation.


**Table 4.** The conversion and selectivity for hydrosilylation of 1-octyne with HMTS in subsequent catalytic runs catalyzed by the same portion of the catalyst **1** isolated from the post-reaction mixture by filtration.

[RC≡CH]:[HSi≡]:[cat] = 1:1.3:2 × 10<sup>−</sup>3; T = 90 ◦C, t = 1 h.

A deterioration of selectivity in subsequent runs was also observed. It can be caused by catalyst modification which occurs with time. Taking into consideration that the concentration of β(E) isomer increases (which according to the literature data is observed for complexes with a bulky steric hindrance), one can suppose that, in subsequent catalytic cycles, due to weakly coordinating properties of imidazolium substituent, binding of reagents or aggregation of the complex occur which results in a larger steric hindrance. However, the possibility of at least fivefold use of the same portion of catalyst significantly reduces the necessary amount of the catalyst and thereby decreases the synthesis costs.

#### **3. Materials and Methods**

#### *3.1. Materials*

All the reagents used in the presented experiments, such as 1-octyne (97%), 1-heptyne (98%), phenylacetylene (98%), 1,1,1,3,5,5,5-heptamethyltrisiloxane (97%), triethylsilane (99%), and n-decane (99%), were supplied by Sigma Aldrich (Pozna ´n, Poland) and used as received. Wilkinson's catalyst and chloro(1,5-cyclooctadiene)rhodium(I) dimer were also purchased from Sigma Aldrich (Pozna ´n, Poland).

#### *3.2. Analytical Techniques*

The yield of the product of hydrosilylation of alkynes was determined by using a Clarus 680 gas chromatograph (Perkin Elmer, Shelton, CT, USA) equipped with a 30 m capillary column Agilent VF-5ms (Santa Clara, CA, USA) and TCD detector, employing the following temperature program: 60 ◦C (3 min.), 10 ◦C min<sup>−</sup>1, and 290 ◦C (5 min.). Characteristic retention times were used for partial identification of the obtained products. The GC-MS analysis was conducted using a Varian 3300 chromatograph (Mundelein, IL, USA) equipped with a 30 m DB-1 capillary column connected to the Finnigan Mat 700 mass detector (Mundelein, IL, USA). For the products obtained, NMR spectra were made with Bruker BioSpin (400 MHz) spectrometer (Ettlingen, Germany) using CDCl3 and CD3CN as a solvent, with tetramethylsilane as the internal standard. Proton chemical shifts are shown in parts per million (δ ppm). The FT-IR in situ measurements were performed using a Mettler Toledo ReactIR 15 instrument (Giessen, Germany). The spectra were recorded with 256 scans for 1 h at 30 s intervals with the resolution of 1 cm<sup>−</sup>1. Changes in the intensity of the bands at 913 cm−<sup>1</sup> and 1600 cm−<sup>1</sup> were recorded using an ATR probe with a diamond window. During the experiments, the decay of the HSi<sup>≡</sup> band originated from HMTS (913 cm<sup>−</sup>1) and the increase in the band originated from the C=C bond (1600 cm−1) that shows the formation of the hydrosilylation reaction product (Figure S41 and Figure S42 in Supplementary Materials) were tracked. CHN elemental analyses were performed on an elemental analyzer Vario EL III (Elementar Analysensysteme GmbH, Langenselbold, Germany).

#### *3.3. Synthesis of Transition-Metal Based Complexes*

Transition-metal based complexes were prepared using the Schlenck technique. In the Schlenck's tube equipped with a magnetic stirrer, a portion of rhodium precursor ([Rh(Cl)(PPh3)3] or [{Rh(μ-Cl)(cod)}2]) with phosphine ligand (3-(4-(diphenylphosphanyl)butyl)-1,2-dimethylimidazolium bromide) was dissolved in toluene. The mixture was stirred for 24 h at room temperature and, after filtration, evaporation of the solvent, and drying under vacuum, the final product was obtained. The detailed synthesis procedures of rhodium ({1,2-dimethyl-3-(diphenylphosphine) butylimidazoliumbromide}bis(triphenylphosphine)chloridorhodium(I) (**1**) and {1,2-dimethyl-3-(diphenylphosphine)butylimidazoliumbromide}(η4-cycloocta-1,5-diene)chloride-rhodium(I)) (**2**) were previously reported by our research group [37]. Spectroscopic characterization and elemental analysis of these complexes are presented in Supplementary Materials.

#### *3.4. General Procedure for Catalytic Tests*

The catalytic activity of Rh phosphine-ligated complexes was tested in the hydrosilylation of 1-heptyne, 1-octyne, and phenylacetylene with 1,1,1,3,5,5,5-heptamethyltrisiloxane (HMTS) or triethylsilane (TES). The hydrosilylation reactions were carried out in the special glass reaction vessel, which was a 5 mL reactor with a side connector for sampling. The reactor was equipped with condenser and magnetic stirrer. In addition, 2 <sup>×</sup> 10−<sup>3</sup> mol of catalyst, 1 mmol of alkyne derivative, 1.3 mmol of HMTS or TES, and 0.5 mmol of n-decane as an internal standard were placed into the reaction vessel. The reaction was carried out at 90 ◦C under air with vigorous stirring for 1 h. Then, the reaction mixture was cooled, centrifuged, and subjected to GC analysis to determine the reaction yield. The obtained retention times for particular isomers were used for partial identification of hydrosilylation products. The products were isolated and subjected to NMR and GC-MS analyses. The obtained spectroscopic data were compared with the literature data [8,12,18,38–43].

#### 1,1,1,3,5,5,5-Heptamethyl-3-[(1Z)Hept-1-Enyl)]Trisiloxane

**1H NMR** (CDCl3)ppm: β (Z): 0.10–0.14 (m, 21H, Si–CH3), 0.92 (t, J = 8.2 Hz, 3H, –CH3), 1.37 (m, 6H, C=C–CH2–(CH2)3), 2.22 (m, 2H, C=C–CH2), 5.34 (dd, J = 14.8, 4.8 Hz, 1H, C=CH–), 6.37 (dt, J = 14.6, 7.4 Hz, 1H, –CH=C), β(E): 5.53 (ddt, J = 21.7, 18.7, 1.5 Hz, 1H, C=CH–), 6.16 (dt, J = 18.7, 6.3 Hz, 1H, –CH=C–). **13C NMR** (CDCl3)ppm: β (Z): 1.81 (–Si–CH3), 13.79 (–CH3), 22.56 (C–CH3), 29.12 (C–C–C–), 31.84 (–C–C–CH3), 33.34 (C=C–C), 127.03 (Si-C=C–), 150.52 (–C=C–), β (E): 127.6 (Si–C=C–), 149.2 (–C=C–). **29Si NMR** (CDCl3)ppm: −35.37 (–OSiCH3), 7.45 (–OSi(CH3)3).

**GC-MS:** β(Z): 305 (3%, [(–OSi(CH3)3)2SiC7H14] <sup>+</sup>), 247.5 (5%, [(–OSi(CH3)3)2SiCH3CH=CH2] +), 229.5 (3%, [–OSi(CH3)3SiCH3C7H14] <sup>+</sup>), 221 (100%, [(-OSi(CH3)3)2SiCH3] <sup>+</sup>), 208.9 (17% [((CH3)3SiO) SiOH(CH3)2SiO]2<sup>+</sup>), β(E): 305 (3%, [(–OSi(CH3)3)2SiC7H14] <sup>+</sup>), 247.5 (7%, [(–OSi(CH3)3)2SiCH3 CH=CH2] <sup>+</sup>), 229.5 (3%, [–OSi(CH3)3SiCH3C7H14] <sup>+</sup>), 221 (100% [(OSi(CH3)3)2SiCH3] <sup>+</sup>), 208.9 (23% [((CH3)3SiO)SiOH(CH3)2SiO]<sup>2</sup>+) α: 305 (2.3%, [(–OSi(CH3)3)2SiCH3C6H12] <sup>+</sup>), 247.5 (5% [(–OSi(CH3)3)2 SiCH3CH=CH2] <sup>+</sup>), 221 (100%, [(–OSi(CH3)3)2SiCH3] <sup>+</sup>), 208.9 (31%, [((CH3)3SiO)SiOH(CH3)2SiO]<sup>+</sup>).

#### 1,1,1,3,5,5,5-Heptamethyl-3-[(1Z)Oct-1-Enyl)]Trisiloxane

**1H NMR** (CDCl3) ppm: β (Z): 0.10–0.15 (m, 21H, Si–CH3), 0.91 (t, J = 6.9 Hz, 3H, –CH3), 1.29–1.45 (m, 8H, C=C–CH2–(CH2)4), 1.95–2.13 (m, 2H, C=C–CH2), 5.37 (dt, J = 14.2, 1.2 Hz, 1H, C=CH–), 6.31 (dt, J = 14.5, 7.4 Hz, 1H, –CH=C), β(E): 5.49 (dt, J = 18.7, 1.5 Hz, 1H, C=CH–), 6.16 (dt, J = 18.7, 6.3 Hz, 1H, –CH=C). **13C NMR** (CDCl3)ppm: 1.71–2.01 (–Si–CH3), 14.35 (–CH3), 22.67 (C–CH3), 29.11, 29.62 (C–C–C–), 31.86 (–C–C–CH3), 36.60 (C=C–C), 126.93 (Si–C=C–), 150.47 (–C=C–). **29Si NMR** (CDCl3)ppm: −35.39 (–OSiCH3), 8.07 (–OSi(CH3)3).

**GC-MS:** β(Z): 317 (25%, [(–OSi(CH3)3)2SiCH=CH(CH2)5CH3)]<sup>+</sup>), 221 (100%, [(–OSi(CH3)3)2SiCH3] +), 208 (19% [((CH3)3SiO)SiOH(CH3)2SiO]<sup>+</sup>), 134 (4%, [–OSi(CH3)3-SiCH3] <sup>2</sup>+), β(E): 317 (30%, [(–OSi(CH3)3)2 SiCH=CH(CH2)5CH3)]<sup>+</sup>), 221 (100%, [(–OSi(CH3)3)2SiCH3] <sup>+</sup>), 208 (20% [((CH3)3SiO)SiOH(CH3)2SiO]<sup>+</sup>), 134 (4%, [–OSi(CH3)3-SiCH3] <sup>2</sup>+), α: 317 (10%, [(–OSi(CH3)3)2SiCH=CH(CH2)5CH3)]<sup>+</sup>), 221 (100%, [(–OSi(CH3)3)2SiCH3] <sup>+</sup>), 208 (18% [(CH3)3SiO)SiOH(CH3)2SiO]<sup>+</sup>), 134 (3.7%, [-OSi(CH3)3-SiCH3] <sup>2</sup>+).

#### 1,1,1,3,5,5,5-Heptamethyl-3-[(1Z)-2-Phenylethenyl]Trisiloxane

**1H NMR** (CD3CN) ppm: β (Z): 0.10–0.24 (m, 21H, –OSiCH3), 5.68 (d, J = 15.5 Hz, 1H, C=CH–Si), 7.31 (d, J = 15.7 Hz, 1H, HC=CH), 7.23–7.39 (m, 3H, C–C=C), 7.49–7.56 (m, 2H, C=C–C), β (E): 6.33 (d, J = 19.3 Hz, 1H, C=CH–Si), 7.04 (d, J = 19.3 Hz, 1H, SiC=CH). **13C NMR** (CD3CN) ppm: −0.5–1.05 (–OSi(CH3)3, –OSiCH3), 127.86 (C=C–C), 128.02, 128.37, 129.37, 139.16, 147.38. **29Si NMR** (CD3CN)ppm: −37.04 (-OSiCH3), 8.07 (–OSi(CH3)3).

**GC-MS**: β(Z): 324 (5%, [(–OSi(CH3)3)2SiCH3–CH=CH2Ph]<sup>+</sup>), 311 (6.5%, [(–OSi(CH3)3)2SiCH= CH2Ph]<sup>+</sup> 221 (18.7% [(–OSi(CH3)3-SiCH3] <sup>+</sup>), 208 (30%, [(–OSi(CH3)3)2Si–]2<sup>+</sup>), 161 (100%, [–OSi(CH3) –Si–CH=CH2–]2<sup>+</sup>), 149 (31%, [–SiCH3–CH=CH2–Ph]2<sup>+</sup>), 104 (5.5%, [CH=CH2–Ph]<sup>+</sup>), 77.1 (6.3%, [–C6H5] +), β(E): 324 (2%, [(–OSi(CH3)3)2SiCH3–CH=CH2Ph]<sup>+</sup>), 311 (5%, [(–OSi(CH3)3)2SiCH=CH2Ph]<sup>+</sup> 221 (100% [(–OSi(CH3)3-SiCH3] <sup>+</sup>), 208 (12%, [(–OSi(CH3)3)2Si–]2<sup>+</sup>), 161 (17%, [–OSi(CH3) –Si–CH=CH2–]2<sup>+</sup>), 149 (15%, [–SiCH3–CH=CH2–Ph]2<sup>+</sup>), 104 (6%, [CH=CH2–Ph]<sup>+</sup>), 77.1 (5%, [–C6H5] +).

#### (Z)-Triethyl(Hept-1-Enyl)Silane

**1H NMR** (CDCl3) ppm: β(Z): 0.63 (m, 4H, –CH2–CH3), 0.94 (m, 11H, –CH3), 1.32 (m, 6H, –CH2–CH2–CH2–), 2.15 (m, 2H, –C=C-CH2–, J=4.8 Hz), 5.42 (dt, J = 14.1, 1.3 Hz 1H, Si–CH=C–), 6.41 (dt, J = 14.4, 7.3 Hz 1H, C=CH–), β(E): 5.59 (dt, J = 18.5, 1.4 Hz, 1H, Si–CH=C–), 6.08 (dt, J = 18.7, 6.3 Hz, 1H, C=CH–). **13C NMR** (CDCl3)ppm: β(Z): 4.73 (Si–CH2–), 7.54 (Si–CH2–CH3), 14.04 (–CH3), 22.61 (C–CH3), 29.49 (–C–C–C–), 31.66 (–C–C–CH3), 34.08 (C=C–C–), 124.89 (Si–C=C–), 150.38 (C=C–C–); β **(E):** 125.52 (Si–C=C–), 148.83 (C=C–C). **29Si NMR** (CDCl3)ppm: −2.83 ((C2H5)3SiH).

**GC-MS**: β(Z): 183 (100%, [–Si(C2H5)2HC=CH2C5H11] <sup>+</sup>), 156 (5%, [SiC2H5HC=CH2C5H11] +), 115.1 (16%, [–Si(C2H5)3] <sup>+</sup>), 99 (37%, [HC=CH2C5H11] <sup>+</sup>), 89.1 (16%, [–Si(C2H5)2] <sup>2</sup>+); β(E): 183 (100%, [–Si(C2H5)2HC=CH2C5H11] <sup>+</sup>), 156 (6%, [SiC2H5HC=CH2C5H11] <sup>+</sup>), 115.1 (14%, [–Si(C2H5)3] <sup>+</sup>), 99 (50%, [HC=CH2C5H11] <sup>+</sup>), 89.1 (28%, [–Si(C2H5)2] <sup>2</sup>+); α**:** 183 (100%, [–Si(C2H5)2HC=CH2C5H11] <sup>+</sup>), 156 (5%, [SiC2H5HC=CH2C5H11] <sup>+</sup>), 115.1 (30%, [–Si(C2H5)3] <sup>+</sup>), 89.1 (70%, [–Si(C2H5)2] <sup>2</sup>+).

#### (Z)-Triethyl(Oct-1-Enyl)Silane

**1H NMR** (CDCl3)ppm: β(Z): 0.55 (m, 4H, –CH2–CH3), 0.96 (t, 3H, –CH3), 0.98 (m, 11H, –CH2–CH3), 1.31–1.46 (m, 8H, –(CH2)4–), 1.96–2.17 (m, 2H, –C=C–CH2–), 5.39 (dd, J = 13.8, 6.1 Hz, 1H, Si–CH=C–), 6.39 (dt, J = 14.4, 7.3 Hz, 1H, C=CH–), β (E): 5.59 (dt, J = 18.7, 1.5 Hz, 1H, –CH=C–), 6.02 (dt, J = 18.7, 6.3 Hz, 1H, C=CH–), α: 5.29 (ddd, J = 12.2, 9.9, 3.8 Hz,1H, –CH=C–). **13C NMR** (CDCl3)ppm: β(Z): 3.55 (Si–CH2–), 7.41 (Si–CH2–CH3), 14.07 (–CH3), 22.63 (C–CH3), 28.80 (–C–C–C–), 29.66 (–C–C–C–), ), 31.38 (–C–C–CH3), 32.78 (C=C–C–), 126.00 (Si–C=C–), 150.36 (C=C–C–), β(E): 125.48 (Si–C=C–), 148.82 (C=C–C–). **29Si NMR** (CDCl3)ppm: −5.78 ((C2H5)3SiH).

**GC-MS:** β(Z): 197 (100%, [SiEt2C8H15] <sup>+</sup>), 141.3 (2% [SiEt3CH=CH]<sup>+</sup>), 115.1 (72.5%, [SiEt3] <sup>+</sup>), 85.1 (6%, [C6H11] <sup>+</sup>), β(E): 197 (100%, [SiEt2C8H15] <sup>+</sup>), 141.3 (15% [SiEt3CH=CH]<sup>+</sup>), 115.1 (15%, [SiEt3] +), 85.1 (26%, [C6H11] <sup>+</sup>), α**:** 197 (100%, [SiEt2C8H15] <sup>+</sup>), 141.3 (12% [SiEt3CH=CH]<sup>+</sup>), 115.1 (30%, [SiEt3] +), 85.1 (25%, [C6H11] +).

#### (E)-Triethyl(Phenyl-1-Ethene)Silane

**1H NMR** (CD3CN) ppm: β(E): (CD3CN) ppm: 0.72–1.04 (m, 15H, –Si(Et)3), 6.50 (d, J = 19.3 Hz, 1H, C=CH–Si), 6.97 (d, J = 19.3 Hz, 1H, HC=C–Si), 7.29 (m, 1H, C=C–C), 7.37 (m, 2H, C=C–C), 7.50 (m, 2H, C–C=C), β(Z): 5.79 (d, J=15.2 Hz, 1H, C=CH–Si) **13C NMR** (CD3CN) ppm: 3.33 (–SiCH2CH3), 7.12 (–SiCH2–), 125.95 (C=C–C–), 126.32, 127.89, 128.48, 138.53, 145.19 **29Si NMR** (CD3CN)ppm: 0.55 (–Si(Et)3).

**GC-MS:** β(E): 189 (100%, [Si(C2H5)2–HC=CH2-Ph], 134.1 (25%, [Si–HC=CH2Ph]3<sup>+</sup>), 115.1 (5%, [–Si(C2H5)3] <sup>+</sup>), 104 (25%, [HC=CH2–Ph]<sup>+</sup>), β(Z): 189 (100%, [Si(C2H5)2–HC=CH2–Ph], 134.1 (52%, [Si–HC=CH2Ph]3<sup>+</sup>), 115.1 (10%, [–Si(C2H5)3] <sup>+</sup>), 104 (30%, [HC=CH2-Ph]<sup>+</sup>).

#### *3.5. General Procedure for Catalyst Isolation and Tests with Subsequent Catalytic Runs*

The catalytic activity of the complex **1** in subsequent catalytic cycles was tested in the hydrosilylation of 1-octyne with 1,1,1,3,5,5,5-heptamethyltrisiloxane (HMTS). The hydrosilylation reaction was conducted in a glass reaction vessel equipped with a condenser and magnetic stirrer. In addition, <sup>2</sup> <sup>×</sup> <sup>10</sup>−<sup>3</sup> mol of catalyst, 1 mmol of 1-octyne, 1.3 mmol of HMTS, and 0.5 mmol of n-decane as an internal standard were placed into the reaction vessel. The reaction was carried out at 90 °C under air with vigorous stirring for 1 h. Then, the reaction mixture was cooled and centrifuged for a few minutes or filtered. Next, the whole amount of product was collected with a syringe equipped with a needle, followed by subjecting to GC analysis to determine the reaction yield. For the catalyst isolated by centrifugation, that remained in the reaction vessel, another portion of the same substrates was added, whereas, for filtered post-reaction mixture, the filter with isolated catalyst was added to the new portion of reagents. In both cases, the isolated catalyst was not subjected to any washing or regeneration. The aforementioned operation was repeated five times, in order to perform five catalytic runs.

#### **4. Conclusions**

The modification of the starting rhodium complexes [Rh(PPh3)3Cl] and [{Rh(μ-Cl)(cod)}2] with the phosphine ligand containing imidazolium ionic liquid as a substituent resulted in obtaining complexes **1** and **2** which are insoluble in the reactants used in the hydrosilylation react on. The above complexes are very durable and stable in air which facilitates their use. Both complexes showed a very high catalytic activity for alkyne hydrosilylation. A complete conversion of alkynes was achieved already after 1 h from the beginning of the reaction. However, the FT-IR in situ analysis of the reactions catalyzed by the complex **1** and the Wilkinson's catalyst has shown that, in the case of the former catalyst, full conversion can be achieved in a considerably shorter time (15 min), whereas, in that of the Wilkinson's catalyst, it requires 55 min. The above studies are a preliminary step to the determination of the kinetics of hydrosilylation of alkynes in the presence of the new rhodium catalysts which we are going to carry out in the immediate future. Three alkynes (1-heptyne, 1-octyne and phenylacetylene) were subjected to hydrosilylation with 1,1,1,3,5,5,5-heptamethyltrisiloxane (HMTS) or triethylsilane (TES). In all the cases, the complexes **1** and **2** were more active than their precursors. In the reactions of hydrosilylation of aliphatic alkynes catalyzed by both complexes, mainly β*(Z)* isomers are formed, which is typical of the major part of rhodium catalysts applied hitherto. In addition, selectivities achieved in this case are higher than those observed for reactions catalyzed by their precursors and range from 89 to 96%. Additionally, a small increase in the selectivity is observed with alkyne chain growth. When comparing selectivities obtained for both catalysts, one can notice that complex **1** is characterized by a higher selectivity. In the reactions with aliphatic alkynes, the effect of the kind of hydrosilylating agent was not observed; both in the reactions with HMTS and TES, *cis* isomer is mainly formed. A somewhat different situation occurs in the hydrosilylation of phenylacetylene, where, depending on the kind of hydrosilylating agent, either *cis* or *trans* isomer predominates. In the reaction with HMTS, the isomer β*(Z)* prevails, whereas, in that with TES, the β*(E)* product is formed. This is caused, besides the steric effect of the phenyl group, also by strongly electronegative nature of triethylsilane whose interaction with phenyl group results in the predominant formation of the *trans* isomer. It is also worth mentioning that the catalyst **2** was active both in reactions with HMTS and TES, whereas its precursor was inactive (maybe because of too mild reaction conditions or too short reaction time).

The most important result of our study is proving the possibility of multiple uses of the same portion of the catalyst. Due to the heterogeneous nature of the complexes **1** and **2**, their isolation and reusing are feasible. The results of conducting five catalytic runs with the use of the same portion of catalyst show that the activity and particularly selectivity are high. It has been proved that the most efficient way of the catalyst isolation from the post-reaction mixture was filtration, which made it possible to reuse the catalyst five times with the preservation of its high activity. Generally, the complexes **1** and **2** that contain the imidazolium- phosphine ligand are characterized by a higher catalytic activity than their precursors. According to the literature data, the rhodium complexes containing ligands with such heteroatoms as P–, N– or NHC– group (which are electron donors) show an increase in the selectivity and yield of the reaction due to a higher stabilization of the Rh-phosphine bond [44–46]. On the other hand, phosphine ligand with imidazolium ionic liquid has a significant effect on the heterogenization of the catalytic system. The obtained catalysts enable their easy isolation and reusability which is of substantial economic and ecological importance.

**Supplementary Materials:** The following are available online at http://www.mdpi.com/2073-4344/10/6/608/s1, Figure S1: 1H NMR spectrum of 1,1,1,3,5,5,5-Heptamethyl-3-[(1Z)hept-1-enyl)]trisiloxane. Figure S2: 13C NMR spectrum of 1,1,1,3,5,5,5-Heptamethyl-3-[(1Z)hept-1-enyl)]trisiloxane. Figure S3: 29Si NMR spectrum of 1,1,1,3,5,5,5-Heptamethyl-3-[(1Z)hept-1-enyl)]trisiloxane. Figure S4: 1H NMR spectrum of 1,1,1,3,5,5,5-Heptamethyl-3-[(1Z)oct-1-enyl)]trisiloxane, Figure S5: 13CNMR spectrum of 1,1,1,3,5,5,5-Heptamethyl -3-[(1Z)oct-1- enyl)]trisiloxane. Figure S6: 29Si NMR spectrum of 1,1,1,3,5,5,5-Heptamethyl-3-[(1Z)oct-1-enyl)] trisiloxane. Figure S7: 1H NMR spectrum of 1,1,1,3,5,5,5-Heptamethyl-3-[(1Z)-2-phenylethenyl]trisiloxane. Figure S8: 13CNMR spectrum of 1,1,1,3,5,5,5-Heptamethyl-3-[(1Z)-2-phenylethenyl]trisiloxane. Figure S9: 29SiNMR spectrum of 1,1,1,3,5,5,5-Heptamethyl-3-[(1Z)-2-phenylethenyl]-trisiloxane, Figure S10: 1H NMR spectrum of (Z)- triethyl(hept-1-enyl)silane. Figure S11: 13C NMR spectrum of (Z)-triethyl(hept-1-enyl)silane, Figure S12: 29SiNMR spectrum of (Z)-triethyl(hept-1-enyl)silane. Figure S13: 1H NMR spectrum of (Z)-triethyl(oct-1-enyl)silane. Figure S14: 13C NMR spectrum of (Z)-triethyl(oct-1-enyl)silane. Figure S15: 29Si NMR spectrum of (Z)-triethyl(oct -1-enyl)silane. Figure S16: 1H NMR spectrum of (E)-triethyl(phenyl-1-ethene)silane. Figure S17: 13C NMR spectrum of (E)-triethyl(phenyl-1-ethene)silane. Figure S18: 29Si NMR spectrum of (E)-triethyl(phenyl-1-ethene)silane. Figure S19: GC chromatogram of 1,1,1,3,5,5,5-Heptamethyl-3-[(1Z)hept-1-enyl)]-trisiloxane, 1,1,1,3,5,5,5-Heptamethyl-3-[(1E)hept -1-enyl)]trisiloxane and 1,1,1,3,5,5,5-Heptamethyl-3-[(α)hept-1-enyl)]trisiloxane. Figure S20: MS spectrum of 1,1,1,3,5,5,5-Heptamethyl-3-[(1Z)hept-1-enyl)]trisiloxane. Figure S21: MS spectrum of 1,1,1,3,5,5,5-Heptamethyl-3-[(1E)hept-1-enyl)]trisiloxane. Figure S22: MS spectrum of 1,1,1,3,5,5,5-Heptamethyl-3-[(1α)hept-1-enyl)]trisiloxane. Figure S23: GC chromatogram of 1,1,1,3,5,5,5-Heptamethyl-3-[(1Z)oct-1-enyl)]trisiloxane1,1,1,3,5,5,5-Heptamethyl -3-[(1E)oct-1-enyl)] trisiloxane and 1,1,1,3,5,5,5-Heptamethyl-3-[(1α)oct-1-enyl)]trisiloxane. Figure S24: MS spectrum of 1,1,1,3,5,5, 5-Heptamethyl-3-[(1Z)oct-1-enyl)]trisiloxane. Figure S25: MS spectrum of 1,1,1,3,5,5,5-Heptamethyl-3- [(1E)oct-1-enyl)]trisiloxane. Figure S26: MS spectrum of 1,1,1,3,5,5,5-Heptamethyl-3-[(1α)oct-1-enyl)]trisiloxane. Figure S27: GC chromatogram of 1,1,1,3,5,5,5-Heptamethyl-3-[(1Z)-2-phenylethenyl]trisiloxane and 1,1,1,3,5,5,5- Heptamethyl-3-[(1E)-2-phenylethenyl]trisiloxane. Figure S28: MS spectrum of 1,1,1,3,5,5,5-Heptamethyl-3-[(1Z)-2 phenylethenyl]trisiloxane. Figure S29: MS spectrum of 1,1,1,3,5,5,5-Heptamethyl-3-[(1E)-2-phenylethenyl]trisiloxane. Figure S30: GC chromatogram of (Z)-triethyl(hept-1-enyl)silane, (E)-triethyl(hept-1-enyl)silane and (α)-triethyl(hept-1-enyl)silane. Figure S31: MS spectrum of (Z)-triethyl(hept-1-enyl)silane. Figure S32: MS spectrum of (E)-triethyl(hept-1-enyl)silane. Figure S33: MS spectrum of (α)-triethyl(hept-1-enyl)silane. Figure S34: GC chromatogram of (Z)-triethyl(oct-1-enyl)silane, (E)-triethyl(oct-1-enyl)silane and (α)-triethyl(oct-1-enyl)silane. Figure S35: MS spectrum of (Z)-triethyl(oct-1-enyl)silane. Figure S36: MS spectrum of (E)-triethyl(oct-1-enyl)silane. Figure S37: MS spectrum of (α)-triethyl(oct-1-enyl)silane. Figure S38: GC chromatogram of (E)-triethyl(phenyl-1-ethene)silane and (Z)-triethyl(phenyl-1-ethene)silane. Figure S39: MS spectrum of (E)-triethyl(phenyl-1-ethene)silane. Figure S40: MS spectrum of (Z)-triethyl(phenyl-1-ethene)silane. Figure S41: FT-IR spectra with characteristic peaks at 1600 cm−<sup>1</sup> and 913 cm−<sup>1</sup> which change with time of the hydrosilylation reaction between 1-octyne and HMTS, carried out in the presence of the Wilkinson's catalyst. Figure S42: FT-IR spectra with characteristic peaks at 1600 cm−<sup>1</sup> and 913 cm−<sup>1</sup> which change with time of the hydrosilylation reaction between 1-octyne and HMTS, carried out in the presence of catalyst **1**.

**Author Contributions:** Catalytic tests, methodology—O.B.; conceptualization—O.B. and H.M.; synthesis of phosphine ligated Rh complexes—M.J.-W.; writing—original draft preparation—O.B. and H.M., writing—review and editing—O.B., M.J.-W. and H.M.; supervision—H.M.; funding acquisition—H.M. and O.B. All authors have read and agreed to the published version of the manuscript.

**Funding:** This research was supported by grant No. POWR.03.02.00-00-I023/17 co-financed by the European Union through the European Social Fund under the Operational Program Knowledge Education Development and grant OPUS UMO-2014/15/B/ST5/04257 funded by National Science Center (Poland).

**Conflicts of Interest:** The authors declare no conflict of interest.

#### **References**

1. Luh, T.-Y.; Liu, S.-T. Synthetic Applications of Allylsilanes and Vinylsilanes. In *The Chemistry in Organic Silicon Compounds*; Rappoport, Z., Apeloig, Y., Eds.; Willey: New York, NY, USA, 1998.


© 2020 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (http://creativecommons.org/licenses/by/4.0/).

## **Piperidinium and Pyrrolidinium Ionic Liquids as Precursors in the Synthesis of New Platinum Catalysts for Hydrosilylation**

**Magdalena Jankowska-Wajda 1,\*, Olga Bartlewicz 1, Przemysław Pietras <sup>2</sup> and Hieronim Maciejewski 1,2**


Received: 11 July 2020; Accepted: 7 August 2020; Published: 10 August 2020

**Abstract:** Six new air-stable anionic platinum complexes were synthesized in simple reactions of piperidinium [BMPip]Cl or pyrrolidinium [BMPyrr]Cl ionic liquids with platinum compounds ([Pt(cod)Cl2] or K2[PtCl6]). All these compounds were subjected to isolation and spectrometric characterization using NMR and ESI-MS techniques. Furthermore, the determination of melting points and thermal stability of the above derivatives was performed with the use of thermogravimetric analysis. The catalytic performance of the synthesized complexes was tested in hydrosilylation of 1-octene and allyl glycidyl ether with 1,1,1,3,5,5,5-heptamethyltrisiloxane. The study has shown that they have high catalytic activity and are insoluble in the reaction medium which enabled them to isolate and reuse them in consecutive catalytic cycles. The most active complex [BMPip]2[PtCl6] makes it possible to conduct at least 10 catalytic runs without losing activity which makes it an attractive alternative not only to commonly used homogeneous catalysts, but also to heterogeneous catalysts for hydrosilylation processes. The activity of the studied catalysts is also affected by the kind of anion and, to some extent, the kind of cation.

**Keywords:** ionic liquids; biphasic catalysis; platinum complexes; hydrosilylation

#### **1. Introduction**

Environmental, economic and technological reasons prompt researchers to pay more and more attention to planning new paths of synthesis of chemical compounds and continuing work on already known reactions and processes successfully employed in the industry, which can be optimized in the aspects of the improvement in yield, reduction in produced waste and possibility of reusing catalysts. Attempts are made at reaching the equilibrium between the costs of conducting a process, its yield, and the environmental impact. All these aspects are in line with the premises of "green chemistry" [1]. One of the processes commonly used on the industrial scale and being the main way of synthesis of organosilicon compounds is hydrosilylation [2–4]. Catalysts for this process are most often transition-metal-based systems of which the most important are platinum complexes, particularly Karstedt ([Pt2{H2C=CHSiMe2)2O}3] and Speier (H2PtCl6/i-PrOH) catalysts [5]. Taking into consideration high price of platinum and the necessity to separate the catalyst from the postreaction mixture (because the presence of heavy metals, even in trace amounts, is impermissible in many applications of the reaction product), efforts are made to employ catalytic systems based on other metals or to heterogenize the most active platinum complexes [5,6].

In recent years, a significant role in the catalytic processes has been played by ionic liquids which can serve as solvents, immobilizing agents for metal complexes, cocatalysts, and catalysts [7–12]. The ionic liquids have been employed in many reactions, one of which is hydrosilylation. In most cases of the latter reaction, ionic liquids dissolved and immobilized metal complexes (mainly those of platinum and rhodium), and formed biphasic systems with the reagents. This role of ionic liquids enabled easy isolation of a catalyst dissolved in them and its reuse. The most often applied ionic liquids were imidazolium [13–17], phosphonium [18,19], ammonium [20], pyrylium [21], and morpholinium [22] ones.

The ionic liquid can also be the element of a complex compound structure which is exemplified by the employment of ionic liquids as substituents in ligands (most often phosphine ones) which at a further stage serve for metal complexation [23–25]. Platinum complexes of this type were applied in reactions of hydrosilylation of alkenes [26] and alkynes [27]. Ionic liquids that contain metal atoms in their structure are another possibility of this kind of application. This group of derivatives is exemplified by halometallate ionic liquids which can be prepared easily by reacting metal halide with organic halide [28–33]. The first and prevalent derivatives have been chloroaluminate ionic liquids [34], albeit currently ionic liquids containing Co, Ir, Au, Ni, Pd, and Pt are also known [28–33]. The platinates known hitherto were solely imidazole derivatives: [EMIM]2[PtCl4], [EMIM]2[PtCl6], and [BMIM]2[PtCl4], [BMIM]2[PtCl6] [35,36]. Recently, our research group obtained next platinates with imidazole and pyridine derivatives and applied them as hydrosilylation catalysts [37,38]. It was the first report on the employment of platinum anionic complexes of this type in the hydrosilylation process. All the complexes have shown high catalytic activity and insolubility in the reagents which enabled them to separate and recycle them.

The simple method of synthesis, high stability, and high activity of the catalysts, as well as the possibility to recycle the latter, which translates into economic and ecological effects, have inspired us to continue studies of this subject and obtain new derivatives. This work was aimed at obtaining new platinates by reactions of platinum salts and platinum chloride complexes with pyrrolidine and piperidine derivatives, their isolation and spectroscopic characterization, as well as the determination of their catalytic activity for hydrosilylation. For the synthesis of the platinates, we have chosen pyrrolidine and piperidine derivatives to compare their properties with those of imidazole and pyridine derivatives obtained earlier [37].

#### **2. Results and Discussion**

To synthesize platinum-containing complexes, two platinum precursors, [Pt(cod)Cl2] and K2[PtCl6], as well as derivatives of piperidine (1-butyl-1-methyl piperidinium chloride, [BMPip]Cl) and pyrrolidine (1-butyl-1-methylpyrrolidinium chloride, [BMPyrr]Cl) have been used. In the starting precursors, platinum was present at different oxidation states which made it possible to obtain tetrachloroplatinates and hexachloroplatinates. In the case of the synthesis of tetrachloroplatinates, depending on the amount of the precursor and ionic liquid, complexes were formed with a different form of the anion. When equimolar amounts of the precursor and ionic liquid were used, the complex with anion in the form of the dimer was obtained, whereas in the case of using two-fold excess of ionic liquid the complex with anion in the monomeric form was created. The syntheses of six new platinum complexes were conducted according to Scheme 1.

**Scheme 1.** The methods of the synthesis of tetrachloroplatinate, hexachlorodiplatinate, and hexachloroplatinate complexes applied in the study.

The synthesis of the above complexes is very simple and consists of the dissolution of platinum precursor and ionic liquid in acetonitrile followed by stirring under reflux for several hours. In the case of the reaction with [Pt(cod)Cl2], after cooling down, the solvent was evaporated together with cyclooctadiene, whereas in the case of the reaction with K2[PtCl6], the filtration of the precipitated KCl preceded the solvent evaporation. All the complexes were obtained with very high yields ranging from 89 to 96%. These compounds are stable in air and no special storage conditions are required.

The obtained complexes were subjected to characterization by 1H and 13C NMR spectroscopy and ESI-MS mass spectrometry. The spectra of the complexes were compared with the spectra of starting reagents. Noteworthy differences were found in the values of chemical shifts of the signals originated from methyl group and CH2 groups bound directly to the nitrogen present in piperidinium and pyrrolidinium cations in the starting ionic liquids compared to tetrachloroplatinate, hexachlorodiplatinate, and hexachloroplatinate complexes with the same cations. The differences in chemical shifts observed for tetrachloroplatinate and hexachloroplatinate complexes range from 0.4 to 0.6 ppm, whereas for hexachlorodiplatinate ones from 0.4 to 0.8 ppm and depend on the kind of cation. The presence of a hydrogen bond had no significant effect on chemical shift values in the 13C NMR spectra.

In the ESI-MS spectra, MS(+) signals from cation and MS(−) ones from chloroplatinate anion were observed. In the MS(+) spectra of all complexes, very intense signals corresponding to molecular peaks of cations: m/z 156.17 [BMPip]<sup>+</sup> and 142.13 [BMPyrr]<sup>+</sup> were present. The MS(−) spectra of tetrachloroplatinate compounds contain multiplets resulting from the presence of three platinum isotopes. The most intense signal was at m/z 335.84 originating from [PtCl4] <sup>2</sup>−. Signals indicating the presence of [PtCl2] <sup>2</sup>−, m/z 265.15 and m/z 300.89, corresponding to [PtCl3] <sup>−</sup>/[Pt2Cl6] <sup>2</sup>−, were also observed. On the other hand, in the case of hexachloroplatinate compounds, signals originating from cations: 156.17 [BMPip]<sup>+</sup> and 142.13 [BMPyrr]<sup>+</sup> were seen. In the MS(−) spectra, characteristic multiplets were visible which originated from platinum isotopes: the most intense of them at m/z 265.15 corresponded to [PtCl2] <sup>2</sup>−, another one at m/z 300.98 was ascribed to [PtCl3] <sup>−</sup>/[Pt2Cl6] <sup>2</sup>−, and the signal at m/z 603.47, which is characteristic of the dimeric complex, originated from [Pt2Cl6] 2−.

For all obtained complexes, melting points were measured and the results are shown in Table 1.


**Table 1.** Melting points of the synthesized chloroplatinate complexes.

The obtained results permit us to conclude that from among complexes with the same cation, hexachloroplatinate complexes have higher melting points than tetrachloroplatinate ones.

Melting points of complexes with anion in the dimeric form are close to or higher than those of hexachloroplatinate complexes. According to the literature, complexes of higher anion symmetry have higher melting points [36] and hexachloroplatinates are characterized by a higher symmetry (Oh) compared to tetrachloroplatinates (D4h symmetry). In the case when complexes with the same anion are compared, one can note that complexes with the [BMPip] cation have higher melting points than complexes with the [BMPyrr] cation.

The thermal stability of the studied compounds was determined by conducting thermogravimetric analysis (TGA) and the results are presented in Figure 1. The temperature at which 10% weight loss occurred was taken as the decomposition temperature (Table 2). The above weight loss value has been chosen to distinguish the decomposition temperature from water desorption temperature. The results listed in Table 2 made it possible to establish that hexachloroplatinate and hexachlorodiplatinate complexes are characterized by higher decomposition temperatures than tetrachloroplatinate complexes. From among the studied complexes, the highest decomposition temperature was found for hexachloroplatinate complex with pyrrolidinium cation, albeit all the temperatures were fairly close one to another and exceeded 200 ◦C which permits to classify all the complexes as thermally stable. This is crucial from the viewpoint of their application in the catalytic processes conducted at elevated temperatures.

**Figure 1.** Thermogravimetric curves of chloroplatinate complexes.


**Table 2.** Decomposition temperatures of the studied compounds at 10% of weight loss.

The essential stage of the study was the application of the prepared complexes as catalysts for hydrosilylation. The activity and selectivity of the developed catalysts were evaluated in the model reactions of hydrosilylation of 1-octene and allyl glycidyl ether with 1,1,1,3,5,5,5-heptamethyltrisiloxane (HMTS). Based on the results of our earlier studies, we conducted the catalytic measurements in analogous conditions, i.e., at 110 ◦C for 1 h [37]. After the reaction completion, the composition of the postreaction mixture was analyzed using GC techniques. The analyses have shown the formation of β-addition products only (in compliance with Scheme 2). No presence of other products, e.g., α-addition or the competitive reaction of olefin isomerization, was observed.

**Scheme 2.** The model reaction of hydrosilylation of 1-octene/allyl glycidyl ether with 1,1,1,3,5,5,5-heptamethyltrisiloxane.

The product yields in the reactions of hydrosilylation of 1-octene and allyl glycidyl ether, catalyzed by the studied platinum complexes are presented in Table 3.


**Table 3.** The product yields in the reactions of hydrosilylation of 1-octene and allyl glycidyl ether with 1,1,1,3,5,5,5-heptamethyltrisiloxane, catalyzed by platinum complexes.

<sup>1</sup> [HSi≡]:[-CH=CH2]:[cat] <sup>=</sup> 1:1:10<sup>−</sup>4; T <sup>=</sup> <sup>110</sup> ◦C; t <sup>=</sup> 1 h, <sup>2</sup> [HSi≡]:[–CH <sup>=</sup> CH2]:[cat] <sup>=</sup> 1:1.2:10<sup>−</sup>4; T <sup>=</sup> <sup>110</sup> ◦C; t <sup>=</sup> 1 h.

Based on the obtained results one can say that all the complexes have shown high catalytic activity and enabled them to obtain a product with high yield. In the case of the reaction with 1-octene, yields obtained in the presence of respective complexes were very similar, whereas, in that of the reaction with allyl glycidyl ether catalyzed by complexes with piperidinium cation, the yield was a bit higher. However, also in the latter case, the differences were small.

#### *Catalysts* **2020**, *10*, 919

All employed complexes are insoluble in reagents which allows their isolation from postreaction mixtures and their reuse in subsequent reaction cycles after adding a new portion of reactants.

The yields obtained in 10 subsequent cycles of the reaction of 1-octene hydrosilylation, catalyzed by the same portion of catalyst, are shown in Figure 2 and Table 4.

**Figure 2.** Yields of the product of hydrosilylation of 1**-**octene with 1,1,1,3,5,5,5-heptamethyltrisiloxane as determined for 10 subsequent reaction cycles catalyzed by the same catalyst portion.

**Table 4.** Product **y**ields and TON values for hydrosilylation of 1-octene with heptamethyltrisiloxane (HMTS) catalyzed by anionic platinum complexes.


[HSi≡]:[CH=CH]:[cat] = 1:1:10<sup>−</sup>4; T = 110 ◦C; t = 1 h.

The obtained results point to significant differences in the activity of the studied platinum-containing catalysts and to a significant influence of the kind of anion on the catalytic activity observed in subsequent reaction cycles. Figure 2 clearly shows that complexes with [PtCl6] 2− and [Pt2Cl6] <sup>2</sup><sup>−</sup> anions (particularly with the former one) show the highest stability and reproducibility. The mentioned complexes are permitted to obtain the product with very high yields in all 10 cycles. The catalytic activity can be easily compared by calculating TON values which are presented in Table 4. It is worth mentioning that the TON values were calculated (for the sake of comparison) for 10 conducted reaction cycles only, although the activity of some complexes was still very high, hence they could be employed in further cycles. Based on the obtained results, one can say that in this case, the effect of the cation is small, albeit the complexes with piperidinium cation have slightly higher activity. From among all complexes studied, the complex [BMPip]2[PtCl6] was the most active. To confirm

the reaction course in subsequent cycles, we studied the reaction using an in situ FTIR probe that made it possible to follow the reaction course in real-time. In the above study, we tracked the decline in the band characteristic of the ≡SiH group. Due to the selective formation of only one product, as shown by chromatograhic analysis of the postreaction mixture, the measured conversion well correlates with the values of product yield determined by chromatographic methods. The obtained ≡SiH conversions in the hydrosilylation reaction catalyzed by the same portion of the [BMPip]2[PtCl6] complex in subsequent cycles are presented in Figure 3.

**Figure 3.** The change in the conversion of ≡SiH as a function of time for the reaction catalyzed by [BMPip]2[PtCl6].

The measurements were conducted for seven cycles and it was found that in each cycle (from the time of the reaction initiation to that at which the final conversion was reached), the reaction course was very fast and lasted from 8 to 10 min. It was also noticed that the inductive period became longer in subsequent cycles, but the reaction profiles were similar and the conversions were on the same level. Analogous measurements were carried out for the reaction catalyzed by the complex [BMPyrr]2[PtCl4] whose activity was the lowest. The obtained results are presented in Figure 4.

In this case, the catalytic activity decreased in subsequent cycles. Although in the first cycle the reaction proceeds very fast (about 10 min), in the further cycles (until reaching the final conversion) the reaction time becomes longer and longer. Moreover, the activity considerably decreases after the fourth cycle (this was also noticed in the results of chromatographic analysis) and the conversion reaches the value of about 30%.

The second reaction studied was hydrosilylation of allyl glycidyl ether in which the activity of all complexes in the first cycle was high (Table 3). This is why we isolated the catalysts and used them in subsequent catalytic cycles. The obtained results are presented in Figure 5 and Table 5. In this case, the results were even more diversified and the highest stability and reproducibility were found for the complexes with a hexachloroplatinic anion.

**Figure 4.** The change in the conversion of ≡SiH as a function of time for the reaction catalyzed by [BMPyrr]2[PtCl4].

**Figure 5.** Yields of the product of hydrosilylation of allyl glycidyl ether with 1,1,1,3,5,5,5-heptamethyltrisiloxane as determined for 10 subsequent reaction cycles catalyzed by the same catalyst portion.


**Table 5.** Product yields and TON values for hydrosilylation of allyl glycidyl ether with HMTS, catalyzed by anionic platinum complexes.

[HSi≡]:[CH=CH]:[cat] = 1:1.2:10<sup>−</sup>4; T = 110 ◦C; t = 1 h.

Additionally, the effect of the cation is more visible because the activity of complexes with piperidinium cation are higher than those with pyrrolidinium cation. This is confirmed by TON values shown in Table 5.

The poorest catalyst turned out to be, also in this case, the complex [BMPyrr]2[PtCl4] whose activity began to decline already in the second cycle thus making it impossible to carry out 10 reaction cycles.

The decline in the activity of tetrachloroplatinic complexes in subsequent cycles of both reactions studied can be explained by gradual leaching of these complexes with new portions of reagents. To confirm this explanation we performed the ICP analysis of postreaction mixtures (obtained after the reaction with 1-octene) for the most stable complex [BMPip]2[PtCl6] and the least stable one [BMPyrr]2[PtCl4]. It has been found that in the case of the former complex, 0.22% of the initial platinum content was leached after the first cycle, whereas after further cycles the amount of leached platinum was below the detection limit. In the case of the latter complex, 2.7% of the initial platinum content was leached after the first cycle and in subsequent cycles the amounts of leached platinum were comparable. The complexes with hexachloroplatinic anions have stronger ionic character than those with tetrachloroplatinic anion, hence they are less susceptible to leaching with reagents. This is more visible in the case of the reaction with allyl glycidyl ether which is a polar compound, hence the decline in the catalytic activity is faster due to a reduction in catalyst concentration caused by stronger leaching. Taking into consideration the effect of the kind of cation, it is worth mentioning that both cations are heterocyclic derivatives with a six-membered ring in the case of piperidine and five-membered ring in that of pyrrolidine. The structure of the six-membered ring additionally stabilizes the complex and due to this, the performance of the complex with pyrrolidinium cation is more reproducible in the reaction with allyl glycidyl ether. In our earlier studies, we determined the activity of chloroplatinic complexes with cations being derivatives of imidazole and pyridine [37]. Considering analogous complexes with piperidinium [BMPip]2[PtCl6] and pyridinium [BMPy]2[PtCl6] cations and comparing TON values for the reaction with octene 96,300 and 95,200, respectively [37] and that with allyl glycidyl ether 88,200 and 60,300, respectively [37], one can additionally confirm that saturated six-membered ring better stabilizes the complex than its aromatic counterpart.

#### **3. Materials and Methods**

#### *3.1. Materials*

All reagents applied in catalytic measurements, i.e., 1-octene, allyl glycidyl ether, n-decane, and 1,1,1,3,5,5,5-heptamethyltrisiloxane were purchased from Sigma Aldrich (Pozna ´n, Poland) and used as received. Additionally, metal precursors: [Pt(cod)Cl2], K2[PtCl6] was supplied by Sigma Aldrich (Pozna ´n, Poland). The ionic liquids: 1-Butyl-1-methylpyrrolidinium chloride [BMPyrr]Cl,1-Butyl-1-methylpiperidinium chloride [BMPip]Cl, were purchased from Iolitec GmbH (Heilbronn, Germany).

#### *3.2. Analytical Techniques*

1H NMR and 13C NMR spectra were recorded in acetonitrile-d3 and chloroform-d, as a solvent, on a Varian 400 operating at 402.6 and 101.2 MHz, respectively. GC analyses were carried out on a Clarus 680 gas chromatograph (Perkin Elmer, Shelton, CT, USA) equipped with a 30 m capillary column Agilent VF-5ms (Santa Clara, CA, USA) and TCD detector, using the temperature program: 60 ◦C (3 min), 10 ◦C min−1, 290 ◦C (5 min). ESI-MS spectra were recorded using a QTOF-type mass spectrometer (Impact HD, Bruker, Ettlingen, Germany). Fourier transform infrared FTIR spectra were recorded on a Bruker Tensor 27 Fourier (Billerica, MA, USA) transform spectrophotometer equipped with a SPECAC Golden Gate, diamond ATR unit with a resolution of 2 cm−1. Thermogravimetric analysis (TGA) was carried out using a TA Instruments TG Q50 analyzer (New Castle, DE, USA) at a linear heating rate of 10 ◦C/min under synthetic air (50 mL/min). The tested samples were placed in a platinum pan and the weight of the samples was kept within 9–10 mg. The experimental error was 0.5% for weight and 1 ◦C for temperature. Melting points were measured on the Melting Point M-565 instrument (Buchi, Essen, Germany) equipped with a video camera. Temperature gradient: 10 ◦C/min. FTIR in situ measurements were performed using a Mettler Toledo ReactIR 15 instrument (Giessen, Germany). For selected samples, spectra were recorded with 256 scans for 1 h at 30 s intervals with the resolution of 1 cm<sup>−</sup>1. Intensity change of the band at 913 cm−1, characteristic of <sup>≡</sup>Si–H bond, was recorded using an ATR probe with a diamond window. The ICP-MS analysis of postreaction samples was carried out on a Perkin Elmer Nexion 300D (Waltham, MA, USA) inductively coupled mass spectrometer.

#### *3.3. Synthesis of Transition-Metal-Based Complexes*

*Synthesis of Bis(1-butyl-1-methylpiperidinium) tetrachloroplatinate(II), [BMPip]2[PtCl4]*

A 25 mL high-pressure Schlenk vessel equipped with a magnetic stirring bar was charged 0.50 g (2.6 mmol) of [BMPip]Cl and 0.50 g (1.3 mmol) of [Pt(cod)Cl2] in hot CH3CN (2 mL). The reaction mixture was stirred for 24 h under reflux. After this time, the solution was cooled to room temperature and the solvent was evaporated. The white product was washed with diethyl ether (3 × 5 mL) and dried under vacuum. Product yield: 93% anal. calcd.

**1H NMR** (ACN-*d6*) δ (ppm): 3.21–3.16 (m, 12H, *J* = 7.24 Hz, N–CH2), 2.88 (s, *J* = 7.14, 6H, N–CH3), 1.90–1.85 (m, 12H, *J* = 7.04, –CH2–), 1.60–1.56 (m, 4H, *J* = 7.08, –CH2–), 1.32–1.26 (m, 4H, *J* = 7.01, –CH2–), 0.91–0.88 (t, 6H, *J* = 7.13, –CH3)

**13C NMR** (ACN-*d6*) δ (ppm): 64.79 (N–CH2), 64.49 (N–CH2), 48.88 (N–CH3), 31.01 (–CH2–), 25.97, 22.12, 19.72 (–CH2–), 13.12 (–CH3)

**ESI-MS(**+**)**: 156.17 [BMPip]<sup>+</sup>

**ESI-MS(**−**)**: 265.15 [PtCl2] <sup>2</sup><sup>−</sup>, 300.86 [PtCl3] <sup>−</sup>/[Pt2Cl6] <sup>2</sup><sup>−</sup>, 335.84 [PtCl4] 2−

*Synthesis of Bis(1-butyl-1-methylpyrrolidinium) tetrachloroplatinate(II), [BMPyrr]2[PtCl4]*

A 25 mL high-pressure Schlenk vessel equipped with a magnetic stirring bar was charged 0.47 g (2.6 mmol) of [BMPyrr]Cl and 0.50 g (1.3 mmol) of [Pt(cod)Cl2] in hot CH3CN (2 mL). The reaction mixture was stirred for 24 h under reflux. After this time, the solution was cooled to room temperature and the solvent was evaporated. The white product was washed with diethyl ether (3 × 5 mL) and dried under vacuum. Product yield: 92% anal. calcd.

**1H NMR** (ACN-*d6*) δ (ppm): 3.49–3.44 (m, 8H *J* = 7.3, –N–CH2), 3.31–3.27 (m, 4H, *J* = 7.4, –N–CH2), 2.99 (s, 6H, *J* = 7.25, –N–CH3), 1.99–1.96 (m, 8H, *J* = 7.21, –CH2–), 1.76–1.73 (m, 4H, *J* = 7.18, –CH2–), 1.42–1.38 (m, 4H, *J* = 7.15, –CH2–), 1.01–0.98 (t, 6H, *J* = 7.13, –CH3)

**13C NMR (**ACN-*d6*) δ (ppm): 64.79 (N–CH2–); 48.63 (–N–CH3); 31.09, 25.95, 21.87, 20.05 (–CH2–) 13.50 (–CH3)

**ESI-MS(**+**)**: 142.13 [BMPyrr]<sup>+</sup>

**ESI-MS(**−**)**: 265.15 [PtCl2] <sup>2</sup><sup>−</sup>, 300.98 [PtCl3] <sup>−</sup>/[Pt2Cl6] <sup>2</sup><sup>−</sup>, 335.86 [PtCl4] 2−

#### *Synthesis of Bis(1-butyl-1-methylpiperidinium) hexachloroplatinate(IV), [BMPip]2[PtCl6]*

A 25 mL high-pressure Schlenk vessel equipped with a magnetic stirring bar was charged 0.37 g (2 mmol) of [BMPip]Cl and 0.53 g (1 mmol) of K2[PtCl6] in hot CH3CN (2 mL) were added. The mixture was stirred for 3 h under reflux followed by filtration by a cannula system. The solvent was evaporated and dried under vacuum to yield orange product. Product yield: 91% anal. calcd.

**1H NMR** (ACN-*d6*) δ (ppm): 3.39–3.29 (m, 12H, *J* = 7.25 Hz, –N–CH2), 3.02 (s, *J* = 7.11, 6H, –N–CH3), 2.06–2.035 (m, 12H, *J* = 7.01, –CH2–), 1.77–1.73 (m, 4H, *J* = 7.06, –CH2–), 1.49–1.43 (m, 4H, *J* = 7.04, –CH2–), 1.08–1.05 (t, 6H, *J* = 7.10, –CH3)

**13C NMR** (ACN-*d6*) δ (ppm): 61.63 (N–CH2), 48.18 (N–CH3), 30.03 (–CH2–), 23.93, 21.19, 20.14, 20.00 (–CH2–), 13.26 (–CH3)

*Synthesis of Bis(1-butyl-1-methylpyrrolidinium) hexachloroplatinate(IV), [BMPyrr]2[PtCl6]*

A 25 mL high-pressure Schlenk vessel equipped with a magnetic stirring bar 0.39 g (2 mmol) of [BMPyrr]Cl and 0.56 g (1.1 mmol) of K2[PtCl6] in hot CH3CN (2 mL) were added. The mixture was stirred for 3 h under reflux followed by filtration by a cannula system. The solvent was evaporated and dried under vacuum to yield orange product. Product yield: 89% anal. calcd.

**1H NMR** (ACN-*d6*) δ (ppm): 3.44–3.43 (m, 8H *J* = 7.11, –N–CH2), 3.28–3.24 (m, 4H, *J* = 7.38, –N–CH2), 2.97 (s, 6H, *J* = 7.21, –N–CH3), 1.80–1.70 (m, 8H, *J* = 7.19, –CH2–), 1.44–1.30 (m, 4H, *J* = 7.17, –CH2–), 1.17–1.13 (m, 4H, *J* = 7.12, –CH2–), 1.01–0.97 (t, 6H, *J* = 7.09, –CH3)

**13C NMR** (ACN-*d6*) δ (ppm): 64.79, 64.53 (N–CH2–); 48.95 (–N–CH3); 31.26, 25.87, 22.07, 19.78 (–CH2–) 13.54 (–CH3)

*Synthesis of Bis(1-butyl-1-methylpiperidinium) hexachlorodiplatinate(II), [BMPip]2[Pt2Cl6]*

A 25 mL high-pressure Schlenk vessel equipped with a magnetic stirring bar was added 0.25 g (1.3 mmol) of [BMPip]Cl and 0.50 g (1.3 mmol) of [Pt(cod)Cl2] in hot CH3CN (2 mL). The mixture was stirred for 24 h under reflux. After this time, the solution was cooled to room temperature and the solvent was evaporated. The white product was washed with diethyl ether (3 × 5 mL) and dried under vacuum. Product yield: 93% anal. calcd.

**1H NMR** (CDCl3) δ (ppm): 3.87–3.81 (m, 4H, *J* = 7.21 Hz, –N–CH2), 3.66–3.60 (m, 8H, *J* = 7.23 Hz, –N–CH2), 3.37 (s, *J* = 7.14, 6H, –N–CH3), 1.91–1.82 (m, 12H, *J* = 7.03, –CH2–), 1.84–1.66 (m, 4H, *J* = 7.06, –CH2–), 1.51–1.41 (m, 4H, *J* = 7.01, –CH2–), 1.03–0.99 (t, 6H, *J* = 7.11, –CH3)

**13C NMR** (CDCl3) δ (ppm): 60.58 (N–CH2), 48.39 (N–CH3), 30.67 (–CH2–), 23.73, 20.52, 19.99, 19.56 (–CH2–), 13.46 (–CH3)

**ESI-MS(**+**)**: 156.17 [BMPip]<sup>+</sup>

**ESI-MS(**−**)**: 265.15 [PtCl2] <sup>2</sup><sup>−</sup>, 300.98 [PtCl3] <sup>−</sup>/[Pt2Cl6] <sup>2</sup><sup>−</sup>, 603.47 [Pt2Cl6] 2

*Synthesis of Bis(1-butyl-1-methylpyrrolidinium) hexachlorodiplatinate(II), [BMPyrr]2[Pt2Cl6]*

A 25 mL high-pressure Schlenk vessel equipped with a magnetic stirring bar was added 0.24 g (1.3 mmol) of [BMPyrr]Cl and 0.50 g (1.3 mmol) of [Pt(cod)Cl2] in hot CH3CN (2 mL). The mixture was stirred for 24 h under reflux. After this time, the solution was cooled to room temperature and the solvent was evaporated. The white product was washed with diethyl ether (3 × 5 mL) and dried under vacuum. Product yield: 90% anal. calcd.

**1H NMR** (CDCl3) δ (ppm): 3.89–3.75 (m, 8H, *J* = 7.42, –N–CH2), 3.63–3.59 (m, 4H, *J* = 7.24, –N–CH2), 3.30 (s, 6H, *J* = 7.20, –N–CH3), 1.81–1.53 (m, 12H, *J* = 7.18, –CH2–), 1.51–1.44 (m, 4H, *J* = 7.16, –CH2–), 1.05–1.01 (t, 6H, *J* = 7.09, –CH3)

**13C NMR** (CDCl3) δ (ppm): 64.51 (N–CH2–), 64.05 (–N–CH3), 30.09, 26.09, 21.63, 19.74 (–CH2–) 13.74 (–CH3)

**ESI-MS(**+**)**: 142.13 [BMPyrr]<sup>+</sup>

**ESI-MS(**−**)**: 265.15 [PtCl2] <sup>2</sup><sup>−</sup>, 300.98 [PtCl3] <sup>−</sup>/[Pt2Cl6] <sup>2</sup><sup>−</sup>, 602.89 [Pt2Cl6] 2−

NMR and ESI-MS spectra of these complexes are presented in Supplementary Materials.

#### *3.4. General Procedure for Catalytic Tests*

To investigate the catalytic activity of platinum anionic complexes the hydrosilylation reactions of 1-octene or allyl glycidyl ether with 1,1,1,3,5,5,5-heptamethyltrisiloxane (HMTS) were carried out. The 5 mL glass reactor equipped with a reflux condenser was charged with 3.68 mmol of 1-octene or 4.41 mmol of allyl glycidyl ether and 3.68 mmol of HMTS. Then, 10−<sup>4</sup> mol of Pt per 1 mol of Si–H was applied. As an internal standard 1 mmol of n-decane was added. The reaction was carried out in the presence of air at 110 ◦C for 1 h, without stirring. After each catalytic cycle, the reaction mixture was cooled down and subjected to GC analysis to determine the reaction yield. The product was isolated and subjected to NMR analyses. Due to a very small amount of catalyst, the products were entirely taken with a needle-equipped syringe after the reaction completion and a new portion of the reaction substrates was added to the reaction vessel, followed by conducting the reaction in the same way as described above. The catalyst remaining in the flask was not washed or regenerated in any way. The above operation was repeated 10 times.

#### *3-octyl-1,1,1,3,5,5,5-heptamethyltrisiloxane:*

**1H NMR** (CDCl3) ppm: 1.36–1.27 (m; 12H; CH2–CH2–CH2); 0.9 (t; 3H; CH2–CH3), 0.48 (t; 2H; Si–CH2), 0.11 (m, 18H, Si– (CH3)3), 0.02 (s, 3H Si–CH3).

**13C NMR** (CDCl3)ppm*:* 33.25 (C–C–C); 31.95 (C–C–C), 29.35, 29.27 (C–C–C), 23.07 (C–CH3), 22.70 (Si–C–C), 17.63 (C–Si), 14.10 (C–CH3), 1.89 (Si–CH3), 0.28 (O–Si–CH3).

**29Si NMR** (CDCl3)ppm: –2.19 (–O–Si–O–), 6.75 (OSi(CH3)3).

*3-(3-glycidyloxypropyl)-1,1,1,3,5,5,5-heptamethyltrisiloxane:*

**1H NMR** (CD3CN)ppm: 3.69 (m, *J* = 17.1 Hz, 1H, –O–CH2–); 3.43 (m, 2H, –CH2–O–CH2-); 3.27 (dd, *J* = 11.5 Hz, 1H, –O–CH2–); 3.08 (m, *J* = 6.8 Hz, 1H, HC–O–CH2–); 2.74 (dd, *J* = 5.1 Hz, 1H, HC–CH2–O); 2.54 (m, *J* = 5.1 Hz, 1H, HC–CH2–O); 0.05 (m, 3H, –SiCH3); 1.59 (m, *J* = 11.3 Hz, 2H, –Si–CH2–CH2–); 0.5 (m, 2H, –Si–CH2–); 0.13 (m, 18H, –Si(CH3)3);

**13C NMR** (CD3CN, δ, ppm): 73.67 (–C–C–O–); 71.41 (–O–C–C–); 50.71 (–C–O–C–); 43.56 (–C–O–C–); 23.16 (–Si–C–C–); 12.77 (–Si–C–); 0.37–1.39 (–Si(CH3)3); –1.0 (–Si–CH3).

**29Si NMR** (CD3CN, δ, ppm): –20.52 (–O–Si–O–), 8.07 (OSi(CH3)3).

NMR spectra of these products are presented in Supplementary Materials.

#### **4. Conclusions**

Six new air-stable anionic platinum complexes were synthesized with high yields in the reaction of a suitable ionic liquid and platinum compound and the complexes were fully characterized. Derivatives of piperidine and pyrrolidine, which were precursors of the formed complexes, have been chosen for the syntheses. Due to high melting points (above 100 ◦C), the newly formed complexes cannot be formally classified into ionic liquids, but the ionic structure significantly influences their catalytic activity and stability. All the complexes proved to be highly active in the reactions of hydrosilylation of 1-octene and allyl glycidyl ether with 1,1,1,3,5,5,5-heptamethyltrisiloxane. Their insolubility (or limited solubility) in the reagents enabled easy isolation from postreaction mixtures and multiple uses of them in subsequent reaction cycles. The performed studies have shown that the kind of anion influences to a large extent the catalytic activity and, first and foremost, the stability of the complexes. The most stable complexes turned to be those with [PtCl6] <sup>2</sup><sup>−</sup> anion due to their stronger ionic character compared to analogous complexes with [PtCl4] <sup>2</sup><sup>−</sup> anion. Moreover, the complexes containing the six-membered heterocyclic ring (piperidinium) are more stable than those containing the five-membered ring (pyrrolidinium). Hexachloroplatinic complexes, particularly the complex [BMPip]2[PtCl6], make an attractive alternative not only to known homogeneous complexes but also to heterogeneous catalysts applied in hydrosilylation processes. High catalytic activity and stability of the latter complex make possible its multiple uses which is of high importance both from an ecological and economic viewpoint.

**Supplementary Materials:** The following are available online at http://www.mdpi.com/2073-4344/10/8/919/s1, Figure S1–Figure S12: NMR Spectra of complexes, Figure S13–Figure S20: ESI-MS spectra of complexes, Figure S21–Figure S26: NMR spectra of isolated products

**Author Contributions:** Synthesis of platinum complexes, methodology, M.J.-W.; catalytic tests, O.B.; FTIR in situ analyzes, O.B. and P.P.; conceptualization, M.J.-W. and H.M.; writing—original draft preparation, M.J.-W. and H.M.; writing—review and editing, M.J.-W., H.M., and O.W.; supervision, H.M.; funding acquisition, H.M. All authors have read and agreed to the published version of the manuscript.

**Funding:** This research was supported by grant OPUS UMO-2014/15/B/ST5/04257, funded by National Science Center (Poland)

**Conflicts of Interest:** The authors declare no conflicts of interest.

#### **References**


© 2020 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (http://creativecommons.org/licenses/by/4.0/).

*Article*
