High Catalytic Efficiency of a Layered Coordination Polymer to Remove Simultaneous Sulfur and Nitrogen Compounds from Fuels

catalysts

Article High Catalytic Eﬃciency of a Layered Coordination Polymer to Remove Simultaneous Sulfur and Nitrogen Compounds from Fuels

Fátima Mirante 1 , Ricardo F. Mendes 2 , Filipe A. Almeida Paz 2,* and Salete S. Balula 1,* 1 REQUIMTE/LAQV & Department of Chemistry and Biochemistry, Faculty of Sciences, University of Porto, 4169-007 Porto, Portugal; [email protected] 2 CICECO-Aveiro Institute of Materials, Department of Chemistry, University of Aveiro, Campus Universitário de Santiago, 3810-193 Aveiro, Portugal; [email protected] * Correspondence: ﬁ[email protected] (F.A.A.P.); [email protected] (S.S.B.)

Received: 1 June 2020; Accepted: 24 June 2020; Published: 2 July 2020

Abstract: An ionic lamellar coordination polymer based on a flexible triphosphonic acid linker, [Gd(H4nmp)(H2O)2]Cl2 H2O (1) (H6nmp stands for nitrilo(trimethylphosphonic) acid), presents high efficiency to remove sulfur and nitrogen pollutant compounds from model diesel. Its oxidative catalytic performance was investigated using single sulfur (1-BT, DBT, 4-MDBT and 4,6-DMDBT, 2350 ppm of S) and nitrogen (indole and quinolone, 400 ppm of N) model diesels and further, using multicomponent S/N model diesel. Different methodologies of preparation followed (microwave, one-pot, hydrothermal) originated small morphological differences that did not influenced the catalytic performance of catalyst. Complete desulfurization and denitrogenation were achieved after 2 h using single model diesels, an ionic liquid as extraction solvent ([BMIM]PF6) and H2O2 as oxidant. Simultaneous desulfurization and denitrogenation processes revealed that the nitrogen compounds are more easily removed from the diesel phase to the [BMIM]PF6 phase and consequently, faster oxidized than the sulfur compounds. The lamellar catalyst showed a high recycle capacity for desulfurization. The reusability of the diesel/H2O2/[BMIM]PF6 system catalyzed by lamellar catalyst was more efficient for denitrogenation than for desulfurization process using a multicomponent model diesel. This behavior is not associated with the catalyst performance but it is mainly due to the saturation of S/N compounds in the extraction phase.

Keywords: layered coordination polymer; oxidative desulfurization; denitrogenation extraction; hydrogen peroxide; lanthanides

1. Introduction One of the main aims of Green chemistry is to minimize the negative impact of the petroleum and chemical industries on the environment and human health. The major sources of air pollution in urban areas are the road fuels [1,2], releasing to the atmosphere diﬀerent pollutants such as carbon monoxide, ammonia, sulfur dioxide (SO2), nitrogen oxides (NOx) and particulate matter. SO2 and NOx emissions can have adverse eﬀects for the environment and for human health [3], with harsh regulations being implemented for sulfur content in road fuels to decrease SO2 emissions [4–6]. In the European Union, strict policies implemented the limit to sulfur level in diesel from 2000 ppm in 1993 to 10 ppm presently. On the other hand, the nitrogen containing levels in fuels are not regulated. NOx emissions results from either the oxidation of nitrogen present compounds in fuels or the oxidation of atmospheric nitrogen at high temperatures [7]. These emissions have already been restrained: the limit for diesel 1 1 powered light duty vehicles decreased from 0.18 g km− for the Euro V standard to 0.08 g km− for Euro VI [8,9].

Catalysts 2020, 10, 731; doi:10.3390/catal10070731 www.mdpi.com/journal/catalysts Catalysts 2020, 10, 731 2 of 15

The industrial process is able to efficiently remove the heterocyclic sulfur and nitrogen compounds (such as benzothiophenes, dibenzothiophenes, thiophenes, quinolines, indoles, carbazoles, acridines, and pyridines, Figure S1 in Electronic Supporting Information) from fuels by hydrotreating processes which require severe conditions (high temperatures >350 ◦C and high pressure 20–130 atm H2)[4,10]. The presence of nitrogen compounds in fuels even in low concentrations (<100 ppm) affects the efficacy of hydrodesulfurization (HDS) reactions since these compounds are strong inhibitors and promoting catalytic deactivation, caused by their competition with sulfur compounds for the active sites of the hydrotreating catalysts [10–13]. Novel technologies capable of complementing or replacing the industrial HDS and hydrodenitrogenation (HDN) processes are needed in order to obtain ultra-low sulfur and nitrogen levels in fuels by more attractive cost-effective methods. In the literature, various examples dedicated to the elimination of sulfur compounds from the fuels are described. The most promising processes are the oxidative desulfurization, combining with the liquid-liquid extraction [4–6]. Most of these processes use hydrogen peroxide as the oxidant, because of the high active oxygen content and with water being the sole by-product. Solid catalysts that allow an easy separation from the reaction media and easy recyclability in consecutive reaction cycles are required. Metal-Organic Frameworks (MOFs) and Coordination Polymers (CPs) comprising organic bridging ligands and metallic centers emerged as promising materials because of their considerable structural diversity and unique properties, such as high porosity, large surface areas and in certain cases high thermal, chemical and hydrolytic stabilities. The investigation and use of these materials as heterogeneous catalysts for desulfurization processes is, however, relatively scarce. Composite materials based on MOFs/CPs have been more regularly applied, employed as host materials to support active homogenous catalysts in their pores or matrices [14–17]. Taking advantage of the possibility of some synthetic modification strategies, the catalytic activity of these materials can be greatly improved [18]. In the last decade, we have been focusing on the design of networks based on polyphosphonic acid ligands and rare-earth metal cations which typically induce the formation of highly robust dense networks [19]. The crystalline hybrid materials designed and developed by our groups typically have a high concentration of acid protons and solvent (typically water) molecules [20]. We are looking to take advantage of this particular structural feature to design better-performing compounds that take advantage of proton mobility and/or exchange. These acid properties can be an important advantage for oxidative desulfurization. In the present work, a positively charged lamellar coordination polymer based on a flexible triphosphonic acid linker, [Gd(H4nmp)(H2O)2]Cl2 H2O(1) [H6nmp stands for nitrilo(trimethylphosphonic) acid], was used as an acid heterogeneous catalyst in oxidative desulfurization and denitrogenation. Environmentally-friendly conditions (low temperature and H2O2/S molar ratio, and an ionic liquid as solvent [BMIM]PF6) were employed. The stability and the recycle capacity of the catalyst was also investigated.

2. Results and Discussion

2.1. Preparation of the Catalyst Our research group have shown the potential of MOFs and CPs as heterogeneous catalysts in various reactions with great industrial interest. The materials tested were based on polyphosphonic acid, self-assembled with various lanthanide cations. The high number of phosphonate groups proved to be important for the catalytic process: the local large concentrations of acidic protons and solvent molecules allowed high conversions rates with low reaction times for reactions such as sulfoxidation of thioanisole [21], conversion of styrene oxide [22] and conversion of benzaldehyde into (dimethoxymethyl)benzene [23]. We further explore the catalytic activity of a positively charged layered 3+ CP obtained by combination of nitrilo(trimethylphosphonic) acid (H6nmp) and Gd ions (Figure1), [Gd(H nmp)(H O) ]Cl 2H O(1). When prepared using a simple one-pot method, this material showed 4 2 2 · 2 a high catalytic activity in four diﬀerent organic reactions: alcoholysis of styrene oxide, acetalization of Catalysts 2020, 10, 731 3 of 15

Catalysts 2020, 10, x FOR PEER REVIEW 3 of 15 benzaldehyde and cyclohexanaldehyde and ketalization of cyclohexanone, with conversion above with95% afterconversion 1–4 h ofabove reaction 95% [24after]. In 1–4 this h of work reaction we further [24]. In explore this work the catalyticwe further activity explore of the this catalytic charged activity2D material of this (charge charged balanced 2D material by chloride (charge ions),balanced in the by desulfurizationchloride ions), in and the denitrogenationdesulfurization and of a denitrogenationmulticomponent of model a multicomponent diesel. Compound model1 diesel.was prepared Compound using 1 threewas prepared diﬀerent methods:using three one-pot different (as methods:reported previously),one-pot (as microwave-assistedreported previously), and microw hydrothermalave-assisted syntheses and hydrothe (Figure1rmal). All syntheses materials present(Figure 1).the All same materials crystalline present phase the with same slight crystalline diﬀerences phase in crystalwith slight morphologies differences and in sizes crystal (Figure morphologies S2 in ESI). andAs previously sizes (Figure reported, S2 in ESI). the use As of previously hydrochloric reported, acid is the crucial use forof thehydrochloric preparation acid of 1is: thecrucial acid for allows the preparationthe protonation of 1 of: the the acid organic allows linker, the retarding protonation the coordinationof the organic process, linker, leadingretarding to thethe formationcoordination of a process,more crystalline leading material.to the formation Not only of that, a more the acidcrystal isline the sourcematerial. ofchloride Not only anions that, the that acid are presentis the source in the ofinterlayer chloride spaces, anions being that responsibleare present forin thethe CPinterl chargeayer balancingspaces, being (we noteresponsible that the usefor ofthe GdCl CP 3chargeas the balancingmetallic source (we note does that not the originate use of GdCl compound3 as the1 ).metallic For additional source does structural not originate details oncompound this compound 1). For additionalwe refer the structural reader to details our past on publicationthis compound [24]. we refer the reader to our past publication [24].

FigureFigure 1. SchematicSchematic representation representation of ofthe the synthesis synthesis and and structural structural features features of of [Gd(H4nmp)(H2O)2]Cl 2H2O(1), with emphasis in the chloride anions present in the interlayer spaces. [Gd(H4nmp)(H2O)2]Cl·2H· 2O (1), with emphasis in the chloride anions present in the interlayer spaces.

Catalysts 2020, 10, 731 4 of 15

2.2. Optimization of Desulfurization Process [Gd(H nmp)(H O) ]Cl 2H O(1) was used as catalyst in the oxidative desulfurization (ODS) of a 4 2 2 · 2 model diesel composed by the most refractory sulfur compounds toward HDS process in liquid fuels, namely 1-benzothiophene (1-BT), dibenzothiophene (DBT), 4-methyldibenzothiophene (4-MDBT) and 4,6-dimethyldibenzothiophene (4,6-DMDBT) in n-octane with a total sulfur concentration of 3 approximately 500 ppm (0.0156 mol dm− ) for each compound (for more detailed information of the experimental conditions the reader is directed to the ESI). Initially catalyst 1, prepared by the three different methods, was tested and their ECODS profiles are presented in Figure2. The ionic liquid 1-butyl-3- methylimidazolium hexafluorophosphate ([BMIM]PF6) was used as extraction solvent. Similar desulfurization efficiency was obtained following different preparation procedures. An initial extraction desulfurization of approximately 40% was achieved after 10 min at 70 ◦C. After the addition of the oxidant the desulfurization increased more appreciably after 40 min and complete desulfurization was found near 2 h. Therefore, the method of catalyst preparation does not seem to have a significant effect in the catalytic efficiency. A small difference is observed between the 40 and 70 min mark and it can be attributed to the overall average particle size. While all methods originated in crystals with plate-like morphology, small differences in size were indeed observed. Contrary to that observed for other materials, the hydrothermal synthesis allows the formation of regular plates with the average size of ca. 10 m. On the other hand, microwave synthesis originated plates as agglomerates ranging from 15 to 30 m. Crystals obtained by the one-pot method presented a more irregular crystal morphology, with sizes varying between 5 and 15 m. Nonetheless, the overall small difference of particle size does not seem to have a direct influence in the catalyst activity itself, achieving the same desulfurization efficiency after 2 h of reaction. Their similar Catalystscrystallinity 2020, 10 may, x FOR indicate PEER REVIEW that the active catalytic centers in the various catalytic samples are identical.5 of 15

100

Desulfurization (%) (%) Desulfurization Microwave 20 Hydrothermal One-pot

0 0 30 60 90 120 150 Time (min)

Figure 2. Desulfurization of a multicomponent model diesel (2350 ppm S) catalyzed by Figure[Gd(H 2.nmp)(H DesulfurizationO) ]Cl 2H ofO( 1a) (20multicomponent mg) prepared bymodel diﬀerent diesel methods (2350 (microwave, ppm S) catalyzed hydrothermal by 4 2 2 · 2 [Gd(H4nmp)(H2O)2]Cl·2H2O (1) (20 mg) prepared by different methods (microwave, hydrothermal and one-pot), using equal volumes of model diesel and [BMIM]PF6 as extraction solvent and H2O2 6 2 2 and(0.64 one-pot), mmol, 30%using aq.) equal at 70 volumes◦C. The of vertical model dashed diesel and line [BMIM]PF indicates the as instant extraction that solvent oxidative and catalytic H O (0.64reaction mmol, was 30% started aq.) byat 70 addition °C. The of vertical oxidant. dashed line indicates the instant that oxidative catalytic reaction was started by addition of oxidant. Because the catalytic activity was virtually the same for 1 obtained by the various synthetic methods, the following optimization of the ECODS system was performed using [Gd(H4nmp)(H2O)2]Cl 2H2O 100 · prepared by the one-pot approach (1op). Various[BMIM]PF6 parameters were investigated, such as extraction solvent, temperature, and oxidant amount,[BMIM]BF4 in order to improve the catalytic performance, 80 solvent free MeCN

40 Desulfurization (%)l (%)l Desulfurization 20

0 0 30 60 90 120 150 Time (min)

Figure 3. Desulfurization of a multicomponent model diesel (2350 ppm S) catalyzed by [Gd(H4nmp)(H2O)2]Cl·2H2O (1op) (20 mg), using different extraction solvents (MeCN,

[BMIM]PF6/[BMIM]BF4) or absence of this solvent, H2O2 (0.64 mmol, 30% aq.) as oxidant at 70 °C. The vertical dashed line indicates the instant that oxidative catalytic reaction was started by addition of oxidant.

Comparing the oxidative desulfurization of each sulfur-based compound, it was possible to notice that after the initial extraction, 1-BT and DBT are more easily transferred from model diesel to the extraction solvent. As 1-BT possesses the smallest molecular size, this transfer is easier, while the presence of methyl substituents at the sterically hindered positions in 4-MDBT and 4,6-DMDBT makes the extraction more challenging [25–27]. During the oxidative catalytic stage, the 1-BT is the most difficult to be oxidized because after 1 h its desulfurization was 67% in contrast to the 92%, 88%

Catalysts 2020, 10, x FOR PEER REVIEW 5 of 15

100 Catalysts 2020, 10, 731 5 of 15

80 the sustainability and cost-efficiency of the process. Three different extraction solvents were investigated: two different ionic liquids (ILs), [BMIM]PF and [BMIM]BF , and the polar organic acetonitrile 60 6 4 (MeCN). Figure3 displays the desulfurization profiles using the various extraction solvents (1:1 model diesel/extraction solvent) and also when no solvent is used. The ECODS system model diesel/[BMIM]PF6 was the40 most efficient, achieving complete desulfurization after 2 h of reaction. In the absence of extraction solvent, practically no oxidative desulfurization occurred, what indicates an

Desulfurization (%) (%) Desulfurization Microwave inefficiency of this layered catalyst in the model diesel phase. Using acetonitrile, an initial extraction of 48% of desulfurization was20 achieved after 10 min. This desulfurizationHydrothermal was not, however, increased during the oxidative desulfurization stage, indicating that the catalystOne-pot is not active using this solvent extraction. When the solvent0 extraction used was the IL [BMIM]BF4, a lower initial extraction was achieved (22%) and the desulfurization0 30 increased 60 after the 90 addition 120 of the oxidant 150 during the first 30 min to 43%, stabilizing this result after this time.Time Therefore, (min) the combination of the catalyst 1op with the [BMIM]PF6 solvent promoted the highest catalytic efficiency, achieving complete desulfurization after only 2 h. ECODS process performed with the 1:1 model diesel/[BMIM]PF system without catalyst Figure 2. Desulfurization of a multicomponent model diesel (2350 ppm6 S) catalyzed by did not resulted any oxidative desulfurization, i.e, after the initial extraction the desulfurization did not [Gd(H4nmp)(H2O)2]Cl·2H2O (1) (20 mg) prepared by different methods (microwave, hydrothermal increased after oxidant addition. This result also indicates that the absorptive capacity of the material is and one-pot), using equal volumes of model diesel and [BMIM]PF6 as extraction solvent and H2O2 negligible.(0.64 mmol, The outstanding 30% aq.) at 70 catalytic °C. The result vertical achieved dashed with line theindicates solvent the [BMIM]PF instant that6 canoxidative be explained catalytic by its immiscibilityreaction was with started the oxidant, by addition creating of oxidant. a three phases system (diesel/H2O2/[BMIM]PF6). This prevents direct contact between the catalyst and the oxidant, avoiding a possible catalyst deactivation.

100 [BMIM]PF6 [BMIM]BF4 80 solvent free MeCN

40 Desulfurization (%)l (%)l Desulfurization 20

0 0 30 60 90 120 150 Time (min) Figure 3. Desulfurization of a multicomponent model diesel (2350 ppm S) catalyzed byFigure [Gd(H 3. nmp)(HDesulfurizationO) ]Cl 2Hof O(a multicomponent1op) (20 mg), usingmodel didieselﬀerent (2350 extraction ppm solventsS) catalyzed (MeCN, by 4 2 2 · 2 [Gd(H4nmp)(H2O)2]Cl·2H2O (1op) (20 mg), using different extraction solvents (MeCN, [BMIM]PF6/[BMIM]BF4) or absence of this solvent, H2O2 (0.64 mmol, 30% aq.) as oxidant at 70 ◦C. The[BMIM]PF vertical6/[BMIM]BF dashed line4) indicatesor absence the ofinstant this solvent, that oxidative H2O2 (0.64 catalytic mmol, reaction30% aq.)was as oxidant started at by 70 addition °C. The ofvertical oxidant. dashed line indicates the instant that oxidative catalytic reaction was started by addition of oxidant. Comparing the oxidative desulfurization of each sulfur-based compound, it was possible to notice that afterComparing the initial the extraction, oxidative 1-BTdesulfurization and DBT are of moreeach sulfur-based easily transferred compound, from model it was diesel possible to the to extractionnotice that solvent.after the Asinitial 1-BT extraction, possesses 1-BT the and smallest DBT molecularare more easily size, transferred this transfer from is easier, model while diesel the to presencethe extraction of methyl solvent. substituents As 1-BT possesses at the sterically the smallest hindered molecular positions size, in 4-MDBT this transfer and 4,6-DMDBTis easier, while makes the thepresence extraction of methyl more challengingsubstituents [25at –the27]. sterically During thehindered oxidative positions catalytic in stage,4-MDBT the 1-BTand 4,6-DMDBT is the most dimakesﬃcult the to beextraction oxidized more because challenging after 1 h its [25–27]. desulfurization During the was oxidative 67% in contrast catalytic to thestage, 92%, the 88% 1-BT and is 80% the ofmost total difficult desulfurization to be oxidized for DBT, because 4-MDBT after and 1 h 4,6-DMDBT,its desulfurization respectively was 67% (Figure in contrast4). The to lower the 92%, electron 88% density of 1-BT compared with the other studied compounds may explain its lower reactivity [27–29]. The studied dibenzothiophene derivative exhibit similar electron densities on the sulfur atom and Catalysts 2020, 10, x FOR PEER REVIEW 6 of 15

Catalystsand 80% 2020 of, total10, x FOR desulfurization PEER REVIEW for DBT, 4-MDBT and 4,6-DMDBT, respectively (Figure 4). The lower6 of 15 electron density of 1-BT compared with the other studied compounds may explain its lower reactivity and[27–29]. 80% Theof total studied desulfurization dibenzothiophene for DBT, derivative 4-MDBT andexhibit 4,6-DMDBT, similar electron respectively densities (Figure on the 4). sulfur The lower atom Catalysts 2020, 10, 731 6 of 15 electronand their density distinct of desulf1-BT comparedurization performancewith the other is studied probably compounds caused the may steric explain hindrance its lower promoted reactivity by [27–29].the methyl The groups. studied dibenzothiophene derivative exhibit similar electron densities on the sulfur atom and their distinct desulfurization performance is probably caused the steric hindrance promoted by their distinct desulfurization performance is probably caused the steric hindrance promoted by the the methyl groups. methyl groups.

Figure 4. Desulfurization results obtained for each sulfur compounds in the model diesel (2350 ppm S)Figure catalyzed 4. Desulfurization by [Gd(H4nmp)(H results2O) obtained2]Cl·2H2O for (1op) each (20 sulfur mg), compounds using [BMIM]PF in the model6, as extraction diesel (2350 solvent, ppm

FigureHS)2O catalyzed2 (0.64 4. Desulfurization mmol, by [Gd(H 30% aq.)nmp)(H results as oxidant obtainedO) ]Cl at 2H70 for °C.O( each 1op su) (20lfur mg), compounds using [BMIM]PF in the model, as extractiondiesel (2350 solvent, ppm 4 2 2 · 2 6 S)H catalyzed2O2 (0.64 mmol,by [Gd(H 30%4nmp)(H aq.) as oxidant2O)2]Cl·2H at 702O ◦(C.1op) (20 mg), using [BMIM]PF6, as extraction solvent, HThe2O2 (0.64effect mmol, of temperature 30% aq.) as oxidant (50, 70 at and70 °C. 80 °C) on the oxidative desulfurization process was analyzedThe and effect results of temperature are summarized (50, 70in Figure and 80 5. ◦HigherC) on temperatures the oxidative were desulfurization not investigated process because was decompositionanalyzedThe effect and results ofof temperatureH are2O2 summarizedcould (50,become 70 in and Figure significant 805 .°C) Higher on above the temperatures oxidative80 °C. The weredesulfurization increase not investigated in processthe reaction because was analyzedtemperaturedecomposition and fromresults ofH 502 areO to2 couldsummarized70 °C become led to inan significant Figure improvemen 5. Higher abovet 80in temperatures ◦theC. Thedesulfurization increase were in not the rate investigated reaction and resulted temperature because in a decompositionsulfur-freefrom 50 to 70model◦ Cof led dieselH to2O an2 productcould improvement becomeafter 2 inh significantof the reac desulfurizationtion. aboveThe desulfurization 80 rate °C. and The resulted increaseprofile in of a sulfur-freeinECODS the reaction at model80 °C temperaturedemonstrateddiesel product from a after rapid 50 2to hincrease 70 of °C reaction. led of todes Theanulfurization improvemen desulfurization aftert in the profilethe first desulfurization 30 of ECODSmin of reaction, at rate 80 ◦ andC and demonstrated resulted after 1 inh ofa a sulfur-freereactionrapid increase 95% model of of total desulfurizationdiesel desulfurization product after after 2 thewas h of first achieved. reac 30tion. min TheComplete of reaction, desulfurization desulfurization and after profile 1 h of was, reactionof ECODS however, 95% at of80 only total °C demonstratedfounddesulfurization after 2 h.a wasrapidTherefore, achieved. increase the Complete ofoptimized desulfurization desulfurization temperatur aftere theis was, 70 first however,°C 30because min only of complete reaction, found afterdesulfurization and 2 h.after Therefore, 1 h of is reactionachievedthe optimized 95% after of 2 temperature totalh. desulfurization is 70 ◦C because was achieved. complete Complete desulfurization desulfurization is achieved was, after however, 2 h. only found after 2 h. Therefore, the optimized temperature is 70 °C because complete desulfurization is achieved after 2 h. 100

100 80

80 60

60 40 70˚C 40 80˚C Desulfurization (%) Desulfurization 20 70˚C 80˚C50˚C Desulfurization (%) Desulfurization 20 0 50˚C 0 306090120 0 0 306090120Time (min)

Figure 5. Desulfurization of a multicomponentTime (min) model diesel (2350 ppm S) catalyzed by [Gd(H nmp)(H O) ]Cl 2H O(1op) (20 mg), using 0.64 mmol of H O oxidant, [BMIM]PF as extraction 4 2 2 · 2 2 2 6 solvent, at diﬀerent temperatures (50, 70, 80 ◦C). The vertical dashed line indicates the instant that oxidative catalytic reaction was started by addition of oxidant. Catalysts 2020, 10, x FOR PEER REVIEW 7 of 15

Figure 5. Desulfurization of a multicomponent model diesel (2350 ppm S) catalyzed by

[Gd(H4nmp)(H2O)2]Cl·2H2O (1op) (20 mg), using 0.64 mmol of H2O2 oxidant, [BMIM]PF6 as extraction solvent, at different temperatures (50, 70, 80 °C). The vertical dashed line indicates the instant that Catalystsoxidative2020, 10 catalytic, 731 reaction was started by addition of oxidant. 7 of 15

The oxidant amount is another important factor because it controls, together with the catalyst, The oxidant amount is another important factor because it controls, together with the catalyst, the oxidative step through oxygen donation. Three different oxidant amounts were used: 75, 50 and the oxidative step through oxygen donation. Three diﬀerent oxidant amounts were used: 75, 50 and 25 µL, corresponding to 0.64, 0.43 and 0.21 mmol, respectively. Results obtained for the ECODS using 25 µL, corresponding to 0.64, 0.43 and 0.21 mmol, respectively. Results obtained for the ECODS using [BMIM]PF6 as extraction solvent at 70 °C, are displayed in Figure 6. Using 0.64 and 0.43 mmol of H2O2 [BMIM]PF6 as extraction solvent at 70 ◦C, are displayed in Figure6. Using 0.64 and 0.43 mmol of H 2O2 a complete desulfurization of the model diesel was obtained after the 2 h. In fact, the desulfurization a complete desulfurization of the model diesel was obtained after the 2 h. In fact, the desulfurization profile after 1 h of reaction is similar using 0.64 and 0.43 mmol of oxidant. On the other hand, using proﬁle after 1 h of reaction is similar using 0.64 and 0.43 mmol of oxidant. On the other hand, using 0.21 mmol of H2O2 an induction period can be observed until 1 h of oxidation catalytic reaction. This 0.21 mmol of H2O2 an induction period can be observed until 1 h of oxidation catalytic reaction. This is is probably due to the lower interaction of the catalytic active centers with the oxidant caused by its probably due to the lower interaction of the catalytic active centers with the oxidant caused by its low low amount. Therefore, the optimized oxidant amount is 0.43 mmol. amount. Therefore, the optimized oxidant amount is 0.43 mmol.

100 0.64 mmol 0.43 mmol 80 0.21 mmol

20 Desulfurization (%) Desulfurization

0 0 306090120 Time (min)

Figure 6. Desulfurization of a multicomponent model diesel (2350 ppm S) catalyzed by Figure 6. Desulfurization of a multicomponent model diesel (2350 ppm S) catalyzed by [Gd(H4nmp)(H2O)2]Cl 2H2O(1op) (20 mg), using different amounts of H2O2 oxidant (0.64, 0.43 [Gd(H4nmp)(H2O)2]Cl·2H· 2O (1op) (20 mg), using different amounts of H2O2 oxidant (0.64, 0.43 and and 0.21 mmol) and [BMIM]PF6 as extraction solvent, at 70 ◦C. The vertical dashed line indicates the 0.21instant mmol) that oxidativeand [BMIM]PF catalytic6 as reaction extraction was solvent, started byat 70 addition °C. Th ofe oxidant.vertical dashed line indicates the instant that oxidative catalytic reaction was started by addition of oxidant. In the next step, the influence of the ratio model diesel/[BMIM]PF6 solvent was studied at 70 ◦C usingIn 0.43 the mmolnext step, of oxidant the influence and 20 mg of the of catalyst. ratio model This optimizationdiesel/[BMIM]PF was6 performed solvent was in studied order to at improve 70 °C usingefficiency 0.43 andmmol sustainability of oxidant ofand the 20 ECODS mg of process. catalyst. Two This di ffoptimizationerent ratios were was performed:performed 1:1in order and 1:0.5 to improve(Figure S3 efficiency in ESI in and the Supportingsustainability Information). of the ECODS The process. volume Two of the different extraction ratios solvent were phase performed: has an 1:1important and 1:0.5 influence (Figure inS3 thein ESI initial in the extraction Supporting step becauseInformation). the sulfur The volume compounds of the are extraction transferred solvent from phasediesel has to this an solventimportant until influence an equilibrium in the pointinitial is extraction achieved thatstep depends because onthe the sulfur nature compounds and amount are of transferredthis solvent. from A higher diesel volume to this ofsolvent [BMIM]PF until6 anmay equilibrium promote a higherpoint is initial achieved extraction that depends of non-oxidized on the naturesulfur compoundsand amount fromof this the solven modelt. dieselA higher to the volume polar phase.of [BMIM]PF Figure6 S3may in ESIpromote shows a thathigher the initial initial extractionextraction of using non-oxidized 1:0.5 of model sulfur diesel compounds/[BMIM]PF from6 wasthe model much lowerdiesel to (8%) the than polar when phase. higher Figure volume S3 in ESIof extraction shows that solvent the initial was extraction used (1:1 using model 1:0.5 diesel of/ model[BMIM]PF diesel/[BMIM]PF6, 31%). After6 was the additionmuch lower of the (8%) oxidant than when(0.43 mmolhigher H volume2O2), the of desulfurization extraction solvent increased was muchused (1:1 faster model when diesel/[BMIM]PF a lower volume of6, extraction31%). After phase the additionwas employed. of the oxidant Using the (0.43 ECODS mmol 1:0.5 H2O model2), the dieseldesulfurization/[BMIM]PF increased6, 99.5% of much desulfurization faster when was a lower found volumeafter 1 h, of instead extraction of 58% phase which was was employed. achieved Using using the a 1:1 ECODS model 1:0.5 diesel model/[BMIM]PF diesel/[BMIM]PF6 system. The6, higher99.5% ofcatalytic desulfurization activity and was consequent found after higher 1 h, desulfurization instead of 58% effi whichciency obtainedwas achieved in the using presence a 1:1 of amodel lower diesel/[BMIM]PFvolume of extraction6 system. solvent, The must higher be related catalytic to the activity lower dispersion and consequent of the solid high catalyster desulfurization in the ECODS efficiencysystem, which obtained can promotein the presence a higher of contact a lower with volu theme aqueous of extraction oxidant. solvent, must be related to the lower dispersion of the solid catalyst in the ECODS system, which can promote a higher contact with the2.3. aqueous Reusability oxidant. Versus Recyclability The stability of the layered coordination polymer [Gd(H nmp)(H O) ]Cl 2H O was investigated 4 2 2 · 2 by its reusability and the recyclability in consecutive ECODS cycles. During all cycles the optimized ECODS conditions were maintained (1:0.5 model diesel/[BMIM]PF6, 0.43 mmol oxidant, at 70 ◦C). Catalysts 2020, 10, x FOR PEER REVIEW 8 of 15

2.3. Reusability Versus Recyclability CatalystsThe2020 stability, 10, 731 of the layered coordination polymer [Gd(H4nmp)(H2O)2]Cl·2H2O was investigated8 of 15 by its reusability and the recyclability in consecutive ECODS cycles. During all cycles the optimized ECODS conditions were maintained (1:0.5 model diesel/[BMIM]PF6, 0.43 mmol oxidant, at 70 °C). TwoTwo didifferentfferent methods methods were were employed employed to recover to recover the catalyst the catalyst aftereach after ECODS each ECODS cycle. In cycle. the “reused” In the method, the model diesel phase was removed and fresh portions of model diesel and H O were added “reused” method, the model diesel phase was removed and fresh portions of model 2diesel2 and H2O2 withoutwere added performing without anyperforming other treatments. any other treatments In the recycled. In the method recycled the method solid catalyst the solid was catalyst separated was fromseparated reaction from medium reaction and medium washed and with washed MeCN with and dried.MeCN Figureand dried.7 presents Figure the 7 resultspresents obtained the results for threeobtained consecutive for three “reused” consecutive and recycled“reused” ECODS and recy cycles.cled ECODS The desulfurization cycles. The desulfurization efficiency decreases efficiency along thedecreases consecutive along reusingthe consecutive cycles. Thisreusing behavior cycles. has This been behavior attributed has been to the attributed saturation to ofthe the saturation extraction of phasethe extraction during the phase reused during process the withreused the process oxidized wi sulfurth the species, oxidized which sulfur prevents species, further which transfer prevents of sulfurfurther compounds transfer of sulfur from compounds the non-polar from phase the tonon-po the ILlar phase, phase decreasingto the IL phase, the systemdecreasing efficiency the system [30]. Onefficiency the other [30]. hand, On the from other the hand, recycling from method, the recycling complete method, desulfurization complete desulfurization was achieved was for theachieved three consecutivefor the three cycles consecutive (Figure cycles7). (Figure 7).

Figure 7. Desulfurization results obtained for three consecutive “reused” and recycled ECODS cycles (1Figure h reaction 7. Desulfurization time), using [Gd(H results4nmp)(H obtained2O) for2]Cl three2H 2consecutiveO(1op) (20 “reused” mg), 0.43 and mmol recycled of H2O ECODS2 oxidant cycles and · 1:0.5(1 h reaction model diesel time),/[BMIM]PF using [Gd(H6, at4nmp)(H 70 ◦C. 2O)2]Cl·2H2O (1op) (20 mg), 0.43 mmol of H2O2 oxidant and

1:0.5 model diesel/[BMIM]PF6, at 70 °C. 2.4. Denitrogenation Process 2.4. DenitrogenationNitrogen compounds Process present in crude oils are predominantly heterocyclic aromatic compounds. and in smaller amounts as non-heterocyclic nitrogen compounds such as aliphatic amines and nitriles. Nitrogen compounds present in crude oils are predominantly heterocyclic aromatic compounds. While the last are easily removed by hydrotreating, heterocyclic nitrogen compounds more difficult and in smaller amounts as non-heterocyclic nitrogen compounds such as aliphatic amines and to remove. These are classified as basic (mainly pyridine derivatives) and neutral (mainly pyrrole nitriles. While the last are easily removed by hydrotreating, heterocyclic nitrogen compounds more derivatives). Quinoline and indole (Figure S1 in ESI) are representative basic and neutral compounds difficult to remove. These are classified as basic (mainly pyridine derivatives) and neutral (mainly that are commonly employed as the components of model fuel oils [10]. Using the optimized ECODS pyrrole derivatives). Quinoline and indole (Figure S1 in ESI) are representative basic and neutral conditions, the oxidative denitrogenation efficiency of the layered [Gd(H nmp)(H O) ]Cl 2H O (1op) compounds that are commonly employed as the components of model4 fuel oils2 2[10].· Using2 the catalyst was studied using a model diesel containing indole (200 ppm of N) and quinoline (200 ppm of optimized ECODS conditions, the oxidative denitrogenation efficiency of the layered N). Figure8 presents the results obtained and it is possible to observe that most of the indole is removed [Gd(H4nmp)(H2O)2]Cl·2H2O (1op) catalyst was studied using a model diesel containing indole (200 from the model diesel during the first 10 min of initial extraction, instead of 80% achieved by quinoline ppm of N) and quinoline (200 ppm of N). Figure 8 presents the results obtained and it is possible to during this first step of denitrogenation process. After the addition of H O oxidant, i.e., during the observe that most of the indole is removed from the model diesel during2 2the first 10 min of initial oxidative denitrogenation stage, the removal of quinoline by its oxidation was rapidly increased and extraction, instead of 80% achieved by quinoline during this first step of denitrogenation process. after 1 h, with both nitrogen compounds being completely extracted from model diesel. The different After the addition of H2O2 oxidant, i.e., during the oxidative denitrogenation stage, the removal of behaviors observed for the two nitrogen compounds can be explained by the interaction between the quinoline by its oxidation was rapidly increased and after 1 h, with both nitrogen compounds being proton/donor molecule indole and the PF anion, which may promote a selective extraction from the completely extracted from model diesel.6− The different behaviors observed for the two nitrogen modelcompounds diesel can [31]. be Quinoline, explained containing by the interaction a six-membered between pyridine the proton/donor ring, the electron molecule lone indole pair on and the the N atom is not part of the aromatic system and extends in the plane of the ring, being responsible for a PF6− anion, which may promote a selective extraction from the model diesel [31]. Quinoline, negativecontaining charge a six-membered on the N atom, pyridine preventing ring, athe lower electron interaction lone pair with on Lewis the N acidic atom IL is [10 not,32 –part34]. of the

Catalysts 2020, 10, x FOR PEER REVIEW 9 of 15 aromatic system and extends in the plane of the ring, being responsible for a negative charge on the NCatalysts atom,2020 preventing, 10, 731 a lower interaction with Lewis acidic IL [10,32–34]. 9 of 15

100

40 Quinoline

Denitrogenation (%) Denitrogenation 20 Indole

0 0 306090120 Time (min)

Figure 8. Denitrogenation profile for the two refractory nitrogen compounds present in diesel Figure(400 ppm 8. Denitrogenation N), catalyzed by [Gd(Hprofile nmp)(Hfor the twoO) ]Clrefractory2H O( 1opnitrogen, 20 mg), compounds 0.43 mmol pr ofesent H O inoxidant diesel (400 and 4 2 2 · 2 2 2 ppm1:0.5 modelN), catalyzed diesel/ [BMIM]PFby [Gd(H4nmp)(H6, at 70 ◦2C.O)2 The]Cl·2H vertical2O (1op dashed, 20 mg), line 0.43 indicates mmol theof H instant2O2 oxidant that oxidativeand 1:0.5 modelcatalytic diesel/[BMIM]PF reaction was started6, at by70 addition°C. The ofvertical oxidant. dashed line indicates the instant that oxidative catalytic reaction was started by addition of oxidant. 2.5. Simultaneous Desulfurization and Denitrogenation Processes 2.5. SimultaneousIn this work, Desulfurization simultaneous and desulfurization Denitrogenation and Processes denitrogenation were also performed with a multicomponentIn this work, model simultaneous diesel containing desulfurization the previous and studieddenitrogenation sulfur compounds were also (1-BT, performed DBT, 4-MDBT with a multicomponentand 4,6-DMDBT, model 2350 ppm diesel of S)containing and indole the (200 prev ppmious ofstudied N) and sulfur quinoline compounds (200 ppm (1-BT, of N). DBT, When 4- MDBTthe desulfurization and 4,6-DMDBT, and denitrogenation2350 ppm of S) and were indole performed (200 ppm separately, of N) and all compoundsquinoline (200 (nitrogen ppm of andN). Whensulfur) the were desulfurization extracted under and thedenitrogenation optimized conditions were performed (20 mg separately, of catalyst, all 0.43 compounds mmol of H (nitrogen2O2 and 1:0.5and sulfur) model were diesel extracted/[BMIM]PF under6, at the 70 ◦optimizedC). However, conditions when denitrogenation (20 mg of catalyst, and 0.43 desulfurization mmol of H2O2 were and 1:0.5performed model simultaneously, diesel/[BMIM]PF practically6, at 70 °C). only However, nitrogen compoundswhen denitrogenation were extracted and fromdesulfurization the model dieselwere performedto the [BMIM]PF simultaneously,6 phase after practically the addition only ofnitr theogen oxidant compounds (Figure S4were in ESI).extracted Complete from removalthe model of dieselN compounds to the [BMIM]PF was also6 phase achieved after after the addition 2 h, instead of the of 1oxidant h obtained (Figure with S4 thein ESI). single Complete denitrogenation removal ofprocess N compounds (Figure8). was These also results achieved may after be related 2 h, instea to thed of use 1 h of obtained an insu ffiwithcient the amount single denitrogenation of H2O2, since a processcompetitive (Figure oxidation 8). These between results N may and be S. related to the use of an insufficient amount of H2O2, since a competitiveOne other oxidation parameter between that canN and contribute S. to the absence of oxidative desulfurization is the low volumeOne of other extraction parameter solvent that to can accommodate contribute bothto the S andabsence N compounds. of oxidative Therefore, desulfurization the simultaneous is the low volumedesulfurization of extraction and denitrogenation solvent to accommodate was performed both S usingand N 0.64 compounds. mmol ofH Therefore,2O2 and equal the simultaneous 1:1 of model desulfurizationdiesel/[BMIM]PF and6 system. denitrogenation An appreciable was performed improvement using of the0.64 desulfurization mmol of H2O2 profileand equal was 1:1 found of model while diesel/[BMIM]PFdenitrogenation profile6 system. was An not appreciable strongly affected, improvement using this of excess the desulfurization of oxidant and theprofile equal was volume found of whileextraction denitrogenation solvent as the profile diesel phasewas not (Figure strongly9). Complete affected, denitrogenationusing this excess was of achievedoxidant and after the 2 h equal (99% volumeafter 1 h) of and extraction for the samesolvent time as thethe desulfurizationdiesel phase (Figure attained 9). 88%Complete (92% after denitrogenation 3 h). These results was achieved indicate afterthat the2 h removal(99% after of sulfur1 h) and is aforffected the same by the time presence the desulfurization of nitrogen, because attained complete 88% (92% desulfurization after 3 h). These was resultsachieved indicate after 2 that h under the removal these experimental of sulfur is conditionsaffected by usingthe presence a single of sulfur nitrogen, model because diesel (Figurecomplete3). desulfurizationThis effect is mainly was noticedachieved during after the2 h oxidativeunder these catalytic experimental stage because conditions the initial using extraction a single ofsulfur each modelcompound diesel was (Figure not significantly 3). This effect affected is mainly by the noticed presence during of S and the Noxidative in the same catalytic model stage diesel. because Figure the 10 initialdepicts extraction the extraction of each for compound each S and was N compound. not signific Itantly is possible affected to by notice the presence that the sulfur of S and compounds N in the samemore model difficult diesel. to desulfurize Figure 10 depicts during the the extraction oxidative fo catalyticr each S stage and N are compound. 1-BT and It the is 4-DMDBT.possible to notice Many thatauthors the sulfur reported compounds that nitrogen more compounds difficult to aredesulfurize significantly during better the extractedoxidative thancatalytic sulfur stage compounds are 1-BT andwhen the ILs 4-DMDBT. are used as Many extraction authors solvents reported [35 ].that Because nitrogen no oxidizedcompounds products are significantly were detected better in theextracted model thandiesel sulfur phase, compounds the oxidative when catalytic ILs are stage used occurs as extrac in majoritytion solvents in the [35]. [BMIM]PF Because6 nophase oxidized and, therefore,products werethe effi detectedciency of in desulfurization the model diesel and denitrogenationphase, the oxid processesative catalytic are strongly stage occurs dependent in majority of the capacity in the [BMIM]PFof S/N extraction.6 phase and, therefore, the efficiency of desulfurization and denitrogenation processes are strongly dependent of the capacity of S/N extraction.

Catalysts 20202020,, 1010,, 731x FOR PEER REVIEW 10 of 15 Catalysts 2020, 10, x FOR PEER REVIEW 10 of 15 100 100

80 80

60 60 (%)

(%) 40 40 Desulfurization 20 Desulfurization 20 Denitrogenation Denitrogenation

Desulfurization or Denitrogenationor Desulfurization 0 Desulfurization or Denitrogenationor Desulfurization 0 0 306090120 0 306090120Time (min) Time (min)

Figure 9. Denitrogenation and desulfurization profile of a model diesel containing approximately 400 Figure 9. Denitrogenation and desulfurization proﬁle of a model diesel containing approximately ppmFigure N 9.and Denitrogenation 2200 ppm of andS, catalyzed desulfurization by [Gd(H profile4nmp)(H of a model2O)2]Cl·2H diesel2O containing(1op) (20 mg),approximately 0.64 mmol 400 of 400 ppm N and 2200 ppm of S, catalyzed by [Gd(H nmp)(H O) ]Cl 2H O(1op) (20 mg), 0.64 mmol of 44 22 2 2 2 2 Hppm2O2 oxidantN and 2200 and 1:1ppm model of S, diesel/[BMIM]PF catalyzed by [Gd(H6, at 70nmp)(H °C. TheO) vertical]Cl·2H· dashO (1op)ed line (20 indicates mg), 0.64 the mmol instant of H2O2 oxidant and 1:1 model diesel/[BMIM]PF6, at 70 ◦C. The vertical dashed line indicates the instant thatH2O oxidative2 oxidant andcatalytic 1:1 model reaction diesel/[BMIM]PF was started by6, additionat 70 °C. ofThe oxidant. vertical dashed line indicates the instant thatthat oxidativeoxidative catalyticcatalytic reactionreaction waswas startedstarted byby additionaddition ofof oxidant.oxidant.

Figure 10. Desulfurization results obtained for each S and N compounds in the model diesel catalyzed by [Gd(HFigure4 10.nmp)(H Desulfurization2O)2]Cl 2H2 resultsO(1op obtained) (20 mg), for using each [BMIM]PF S and N compounds6, as extraction in the solvent, model H diesel2O2 (0.64 catalyzed mmol, Figure 10. Desulfurization· results obtained for each S and N compounds in the model diesel catalyzed 30%by [Gd(H aq.) as4nmp)(H oxidant,2O) at2]Cl·2H 70 ◦C. 2O (1op) (20 mg), using [BMIM]PF6, as extraction solvent, H2O2 (0.64 mmol,by [Gd(H 30%4nmp)(H aq.) as oxidant,2O)2]Cl·2H at 270O °C.(1op) (20 mg), using [BMIM]PF6, as extraction solvent, H2O2 (0.64 Themmol, catalytic 30% aq.) easffi oxidant,ciency ofat 70 [Gd(H °C. nmp)(H O) ]Cl 2H O/[BMIM]PF system was studied for 4 2 2 · 2 6 variousThe cycles catalytic of simultaneous efficiency of desulfurization [Gd(H4nmp)(H and2O)2 denitrogenation]Cl·2H2O/[BMIM]PF of multicomponent6 system was studied S/N model for diesel.variousThe This cycles catalytic method of simultaneous wasefficiency chosen of insteaddesulfurization [Gd(H of4nmp)(H the recycling and2O) 2denitrogenation]Cl·2H process2O/[BMIM]PF to face sustainabilityof multicomponent6 systemand was cost studiedS/N effectivity model for thatdiesel.various are This importantcycles method of simultaneous topics was to chosen consider desulfurization instead for industrial of the and re application. cyclingdenitrogenation process toof facemulticomponent sustainability S/N and model cost effectivitydiesel. This that method are important was chosen topics instead to consider of the for re industrialcycling process application. to face sustainability and cost 2.6.effectivity Reusability that Studiesare important topics to consider for industrial application. 2.6. ReusabilityReusability Studies experiments were performed using an excess of 0.64 mmol of oxidant and a ratio 2.6. Reusability Studies of 1:1Reusability of model diesel experiments/[BMIM]PF were6. performed Figure 11 displays using an the excess results of 0.64 obtained mmol for of theoxidant three and consecutive a ratio of cycles1:1 ofReusability model presenting diesel/[BMIM]PF experiments the data obtained 6were. Figure performed after 11 displays the initial using the extraction an results excess obtained (before of 0.64 oxidantformmol the ofthree addition) oxidant consecutive and and a after ratio cycles 2 of h ofpresenting1:1 oxidative of model the catalyticdiesel/[BMIM]PF data obtained stage. The 6after. Figure initial the 11 extractioninitial displays extraction isthe not results a ff(beforeected obtained byoxidant the for three theaddition) three reusing consecutive and cycles. after For 2cycles h the of denitrogenationoxidativepresenting catalytic the data process stage. obtained theThe transfer afterinitial the extraction of initial N compounds extraction is not affected to (before the extraction by oxidant the three phaseaddition) reusing still increasedand cycles. after For 2 along hthe of thedenitrogenationoxidative cycles (86,catalytic 93, process 97% stage. for the The the transfer 1st,initial 2nd of extraction andN compounds 3rd cycles,is not to respectively).affected the extraction by the The phasethree oxidative stillreusing increased denitrogenation cycles. along For thethe ecyclesdenitrogenationfficiency (86, is93, maintained 97% process for the throughoutthe 1st, transfer2nd and 3 theofrd cycles,N various compounds respectively). cycles to and the completeTheextraction oxidative denitrogenation phase denitrogenation still increased was efficiency achievedalong the st nd rd afteriscycles maintained 2 (86, h. The93, 97%throughout same for is the not, 1 the however,, 2 various and 3 observed cycles cycles, and respectively). for complete the oxidative denitrogenationThe oxidative desulfurization, denitrogenation was achieved where a efficiencyafter decrease 2 h. Theis maintained same is not, throughout however, observedthe various for cycles the oxidat and ivecomplete desulfurization, denitrogenation where wasa decrease achieved of efficiency after 2 h. The same is not, however, observed for the oxidative desulfurization, where a decrease of efficiency

Catalysts 2020, 10, 731 11 of 15 Catalysts 2020, 10, x FOR PEER REVIEW 11 of 15 of efficiency was found from the 1st to the 2nd and the 3rd cycles: 88, 62 and 54%, respectively. was found from the 1st to the 2nd and the 3rd cycles: 88, 62 and 54%, respectively. The efficiency of the The efficiency of the desulfurization process is, therefore, diminished by the presence of N compounds desulfurization process is, therefore, diminished by the presence of N compounds during the various during the various consecutive cycles. The higher difficulty of desulfurization process in the presence consecutive cycles. The higher difficulty of desulfurization process in the presence of N compounds of N compounds must be even more pronounced during the reusing cycles because the extraction must be even more pronounced during the reusing cycles because the extraction phase became more phase became more saturated with N and S compounds from the 1st to the 2nd and to the 3rd cycles. saturated with N and S compounds from the 1st to the 2nd and to the 3rd cycles. Desulfurization is Desulfurization is more sensitive to deactivation during reusing because nitrogen compounds are more sensitive to deactivation during reusing because nitrogen compounds are significantly better significantly better extracted than sulfur ones using IL as extraction solvents [35]. extracted than sulfur ones using IL as extraction solvents [35].

Figure 11. Desulfurization and denitrogenation results obtained for three consecutive reused cycles (initial extraction and after 2 h of catalytic oxidation), using [Gd(H nmp)(H O) ]Cl 2H O(1op) (20 mg), Figure 11. Desulfurization and denitrogenation results obtained4 for three2 2 consecutive· 2 reused cycles 0.64(initial mmol extraction of H2O2 oxidant and after and 2 1:1h of model catalytic diesel oxidation),/[BMIM]PF using6 system, [Gd(H at4 70nmp)(H◦C. 2O)2]Cl·2H2O (1op) (20 2.7. Structuralmg), 0.64 Stability mmol of of H [Gd(H2O2 oxidantnmp)(H and O)1:1 model]Cl 2H diesel/[BMIM]PFO 6 system, at 70 °C. 4 2 2 · 2 2.7.The Structural chemical Stability robustness of [Gd(H and structural4nmp)(H2O) stability2]Cl·2H of2O the catalyst was evaluated by performing several characterizations techniques of the solid after the ODS and ODN processes. As depicted in Figure 12, The chemical robustness and structural stability of the catalyst was evaluated by performing [Gd(H4nmp)(H O) ]Cl 2H O (1) suffers a single-crystal-to-single-crystal transformation under these several characterizations2 2 · 2 techniques of the solid after the ODS and ODN processes. As depicted in catalytic conditions. Remarkably, 1 withstands the presence of H2O2. Nevertheless, high temperatures Figure 12, [Gd(H4nmp)(H2O)2]Cl·2H2O (1) suffers a single-crystal-to-single-crystal transformation led to the complete transformation of 1, even after the first cycle. [Gd(H4nmp)(H2O)2]Cl 2H2O (1) under these catalytic conditions. Remarkably, 1 withstands the presence of H2O2. Nevertheless,· high low stability under harsh conditions was previously reported by us. [36] The catalytic structure 1 temperatures led to the complete transformation of 1, even after the first cycle. suffers a single-crystal-to-single-crystal transformation at high temperatures and relative humidity, [Gd(H4nmp)(H2O)2]Cl·2H2O (1) low stability under harsh conditions was previously reported by resulting in a different 2D layered material. In this case, the combination of high temperature and us.[36] The catalytic structure 1 suffers a single-crystal-to-single-crystal transformation at high the presence of the IL led to another of such structural transformation which resulted in yet another temperatures and relative humidity, resulting in a different 2D layered material. In this case, the distinct, unknown phase. combination of high temperature and the presence of the IL led to another of such structural transformation which resulted in yet another distinct, unknown phase. Although the structural elucidation of the catalytic material using conventional powder X-ray analysis was not possible, the remaining solid state characterization allowed us to draw some conclusions (see Section 3 in ESI document for further details). The transformation is accompanied with the release of the chloride anion (possibly in the form of hydrochloric acid, very much as recently observed by us). This release is evident by the elemental distribution analysis performed by energy dispersive X-ray spectroscopy (EDS) in the transformed material (Figure S5 in ESI), and supported by previous transformation of this material under different conditions [36]. After the catalytic process, though the particles appear contaminated with sulfur and fluoride species (which can be attributed to residual IL), the quantification of the Gd and P elements could still be performed. This led to a P:Gd ratio of 3:1 after the catalytic use (maintaining the one Gd3+ center per each H6nmp linker), which is the same as determined for the initial layered structure. Therefore, we deduce that the observed structural transformation was not accompanied with the loss of either of the basic building units of the compound. In addition, FT-IR analysis (Figure S6 in the Supporting Information)

Catalysts 2020, 10, x FOR PEER REVIEW 12 of 15 reveals the presence of virtually the same vibrational bands as for 1, characteristic of phosphonate- based materials, suggesting that no significant changes in coordination modes were observed. While the crystalline structure could not unveiled, a Pawley refinement was performed (data not shown). The calculated new unit cell (cell parameters: a = 7.91 Å, b = 10.33 Å, c = 11.59 Å; α = 103.95°, β = 95.46° and γ = 90.64°) suggests a change to a less symmetric cell (from Ia to possibly P-1), with a decrease more accentuated in one of the cell axis (from 17.48 to 10.33 Å). Because 1 is a 2D layered material, the release of the chlorine anions (that are present in the interstitial space acting as charge-balancing anions) might lead to a significant decrease in the interlayer distance, resulting in registered variations of the unit cell parameters. Once again, this structural behavior agrees well with the previously reported transformation of 1 [36,37]. While the powder X-ray diffraction data confirm that the crystalline structure of layered [Gd(H4nmp)(H2O)2]Cl·2H2O is modified after one catalytic cycle, we emphasize that the new Catalystsstructure2020 remained, 10, 731 active and without structural change after three consecutive ECODS 12cycles of 15 (Figure 12). The new compound retains the same plate-like crystal morphology.

Figure 12. Powder X-ray diffraction patterns of the layered [Gd(H nmp)(H O) ]Cl 2H O, before and 4 2 2 · 2 afterFigure the 12. 1st Powder and 3rd X-ray ECODS diffraction cycles. patterns of the layered [Gd(H4nmp)(H2O)2]Cl·2H2O, before and after the 1st and 3rd ECODS cycles. Although the structural elucidation of the catalytic material using conventional powder X-ray analysis3. Conclusions was not possible, the remaining solid state characterization allowed us to draw some conclusionsThe ionic (see Sectionlamellar3 incoordination ESI document polymer for further based details). on a Theflexible transformation triphosphonic is accompanied acid linker, with the release of the chloride anion (possibly in the form of hydrochloric acid, very much as recently [Gd(H4nmp)(H2O)2]Cl2 H2O (H6nmp stands for fornitrilo(trimethylphosphonic) acid) presented a observedhigh catalytic by us). efficiency This release to oxidize is evident the most by therefrac elementaltory sulfur distribution and nitrogen analysis compounds performed present by energy in real dispersivediesel (mainly X-ray dibenzothiophene spectroscopy (EDS) derivative, in the transformed indole, and material quinolone). (Figure The S5 different in ESI), andmethods supported followed by previousfor its preparation transformation (microwave, of this materialone-pot, underhydrothermal) different originated conditions some [36]. Aftermorphological the catalytic differences, process, thoughas the size the particlesand shape appear of obtained contaminated particles; with however, sulfur andthis fluoridedid not influence species (which its catalytic can be performance. attributed to residual IL), the quantification of the Gd and P elements could still be performed. This led to a P:Gd Using the model diesel/H2O2/[BMIM]PF6 system, complete desulfurization and denitrogenation were 3+ ratioachieved of 3:1 after after 2 theh of catalytic reaction use using (maintaining sulfur or thenitrogen one Gd modelcenter diesel, per respectively. each H6nmp The linker), ionicwhich liquid is the same as determined for the initial layered structure. Therefore, we deduce that the observed [BMIM]PF6 was the extraction solvent selected since its efficiency was higher than MeCN and other structuralmore hydrophilic transformation ionic liquids. was not When accompanied the single with model the diesel loss of was either replaced of the by basic the building multicomponent units of theS/N compound. model diesel, In addition,the desulfurization FT-IR analysis efficiency (Figure decreases S6 in the from Supporting 100% to Information)88% after 2 h, reveals while thethe presencedenitrogenation of virtually effectivity the same was vibrational maintained. bands The as initia for 1,l characteristicextraction (before of phosphonate-based oxidant addition) materials,for sulfur suggestingand nitrogen that was no significantmaintained changes when insingle coordination model diesels modes were were replaced observed. by the multicomponent While the crystalline structure could not unveiled, a Pawley refinement was performed (data not shown). The calculated new unit cell (cell parameters: a = 7.91 Å, b = 10.33 Å, c = 11.59 Å; α = 103.95◦, β = 95.46◦ and γ = 90.64◦) suggests a change to a less symmetric cell (from Ia to possibly P-1), with a decrease more accentuated in one of the cell axis (from 17.48 to 10.33 Å). Because 1 is a 2D layered material, the release of the chlorine anions (that are present in the interstitial space acting as charge-balancing anions) might lead to a significant decrease in the interlayer distance, resulting in registered variations of the unit cell parameters. Once again, this structural behavior agrees well with the previously reported transformation of 1 [36,37]. While the powder X-ray diffraction data confirm that the crystalline structure of layered [Gd(H nmp)(H O) ]Cl 2H O is modified after one catalytic cycle, we emphasize that the new structure 4 2 2 · 2 remained active and without structural change after three consecutive ECODS cycles (Figure 12). The new compound retains the same plate-like crystal morphology. Catalysts 2020, 10, 731 13 of 15

3. Conclusions The ionic lamellar coordination polymer based on a flexible triphosphonic acid linker, [Gd(H4nmp)(H2O)2]Cl2 H2O (H6nmp stands for fornitrilo(trimethylphosphonic) acid) presented a high catalytic efficiency to oxidize the most refractory sulfur and nitrogen compounds present in real diesel (mainly dibenzothiophene derivative, indole, and quinolone). The different methods followed for its preparation (microwave, one-pot, hydrothermal) originated some morphological differences, as the size and shape of obtained particles; however, this did not influence its catalytic performance. Using the model diesel/H2O2/[BMIM]PF6 system, complete desulfurization and denitrogenation were achieved after 2 h of reaction using sulfur or nitrogen model diesel, respectively. The ionic liquid [BMIM]PF6 was the extraction solvent selected since its efficiency was higher than MeCN and other more hydrophilic ionic liquids. When the single model diesel was replaced by the multicomponent S/N model diesel, the desulfurization efficiency decreases from 100% to 88% after 2 h, while the denitrogenation effectivity was maintained. The initial extraction (before oxidant addition) for sulfur and nitrogen was maintained when single model diesels were replaced by the multicomponent diesel. However, the extraction of nitrogen compounds is higher (86%) than the sulfur compounds (36%), what contribute largely for the higher efficiency of denitrogenation process. The recycle capacity of the lamellar catalyst was studied for consecutive desulfurization processes and the catalytic efficiency was maintained between cycles. This result indicates that [Gd(H4nmp)(H2O)2]Cl2 H2O is a stable catalyst, although some structural adjustment occurred to form the active heterogeneous catalyst. An improvement in the reuse capacity of the diesel/H2O2/[BMIM]PF6 system need to be performed in the near future, since the desulfurization process catalyzed by the lamellar material loss efficiency in consecutive ECODS cycles, probably caused by the saturation of the extraction [BMIM]PF6 phase with sulfur and nitrogen compounds.

Supplementary Materials: The following are available online at http://www.mdpi.com/2073-4344/10/7/731/s1, Figure S1: The representative sulfur and nitrogen compounds used in this work to prepare model diesels. 1-benzothiophene (1-BT), dibenzothiophene (DBT), 4-methyldibenzothiophene (4-MDBT) and 4,6-dimethyldibenzothiophene (4,6-DMDBT), Figure S2: Powder X-ray diffraction and SEM images of [Gd(H4nmp)(H2O)2]Cl 2H2O (1) obtained using different experimental methods (op—one-pot; mw—Microwave Assisted; ht—Hydrothermal),· Figure S3: Desulfurization of a multicomponent model diesel (2350 ppm S) catalyzed by layered MOF [Gd(H nmp)(H O) ]Cl 2H O (20 mg), using 0.43 mmol of H O oxidant and different volume of 4 2 2 · 2 2 2 [BMIM]PF6 extraction solvent (1:1 and 1:0.5 model diesel/[BMIM]PF6), at 70 ◦C. The vertical dashed line indicates the instant that oxidative catalytic reaction was started by addition of oxidant, Figure S4: Denitrogenation and desulfurization profile of a model diesel containing approximately 400 ppm N and 2200 ppm of S, catalyzed by layered MOF [Gd(H nmp)(H O) ]Cl 2H O (20 mg), 0.43 mmol of H O oxidant and 1:0.5 model diesel/[BMIM]PF6, 4 2 2 · 2 2 2 at 70 ◦C. The vertical dashed line indicates the instant that oxidative catalytic reaction was started by addition of oxidant, Figure S5: SEM, mapping and EDS spectra of compound [Gd(H nmp)(H O) ]Cl 2H O (1) after catalytic 4 2 2 · 2 use for one cycle of ECODS, with a P:Gd ratio of 3:1, Figure S6: FT-IR spectra of layered [Gd(H4nmp)(H2O)2]Cl 2H2O before and after catalytic use for one ECODS cycle. · Author Contributions: Conceptualization, F.A.A.P and S.S.B.; Data curation, F.M. and R.F.M.; Formal analysis, F.M. and R.F.M.; Funding acquisition, S.S.B.; Investigation, F.M.; Project administration, S.S.B.; Supervision, S.S.B. and F.A.A.P.; Validation, F.A.A.P. and R.F.M.; Visualization, F.M.; Writing—original draft, S.S.B.; Writing—review & editing, F.A.A.P and S.S.B. All authors have read and agreed to the published version of the manuscript. Funding: This work was partly funded through the project REQUIMTE-LAQV (Ref. UID/QUI/50006/2019) and the project GlyGold, PTDC/CTM-CTM/31983/2017, and partially developed within the scope of the project CICECO-Aveiro Institute of Materials, UIDB/50011/2020 & UIDP/50011/2020, LAQV-REQUIMTE (UIDB/50006/2020) and CQE (UIDB/00100/2020) research units, financed by national funds through the FCT/MCTES (Fundação para a Ciência e a Tecnologia / Ministério da Ciência, Tecnologia e Ensino Superior) and when appropriate co-financed by FEDER (Fundo Europeu de Desenvolvimento Regional) under the PT2020 Partnership Agreement. FCT is also gratefully acknowledged for the Junior Research Position CEECIND/00553/2017 (to RFM). Conflicts of Interest: The authors declare no conflict of interest. Catalysts 2020, 10, 731 14 of 15

References

1. Chossière, G.P.; Malina, R.; Allroggen, F.; Eastham, S.D.; Speth, R.L.; Barrett, S.R.H. Country- and manufacturer-level attribution of air quality impacts due to excess NOx emissions from diesel passenger vehicles in Europe. Atmos. Environ. 2018, 189, 89–97. [CrossRef] 2. Frey, H.C. Trends in onroad transportation energy and emissions. J. Air Waste Manag. Assoc. 2018, 68, 514–563. [CrossRef][PubMed] 3. Koolen, C.D.; Rothenberg, G. Air Pollution in Europe. ChemSusChem 2019, 12, 164–172. [CrossRef][PubMed] 4. Stanislaus, A.; Marafi, A.; Rana, M.S. Recent advances in the science and technology of ultra low sulfur diesel (ULSD) production. Catal. Today 2010, 153, 1–68. [CrossRef] 5. Chandra Srivastava, V. An evaluation of desulfurization technologies for sulfur removal from liquid fuels. RSC Adv. 2012, 2, 759–783. [CrossRef] 6. Houda, S.; Lancelot, C.; Blanchard, P.; Poinel, L.; Lamonier, C. Oxidative Desulfurization of Heavy Oils with High Sulfur Content: A Review. Catalysts 2018, 8, 344. [CrossRef] 7. Re¸sito˘glu, I.A.;˙ Altini¸sik,K.; Keskin, A. The pollutant emissions from diesel-engine vehicles and exhaust aftertreatment systems. Clean Technol. Environ. Policy Vol. 2015, 17, 15–27. [CrossRef] 8. Roy, S.; Baiker, A. NOx Storage Reduction Catalysis: From Mechanism and Materials Properties to − Storage Reduction Performance. Chem. Rev. 2009, 109, 4054–4091. [CrossRef] − 9. Yang, L.; Franco, V.; Mock, P.; Kolke, R.; Zhang, S.; Wu, Y.; German, J. Experimental Assessment of NOx Emissions from 73 Euro 6 Diesel Passenger Cars. Environ. Sci. Technol. 2015, 49, 14409–14415. [CrossRef] 10. Prado, G.H.C.; Rao, Y.; de Klerk, A. Nitrogen Removal from Oil: A Review. Energy Fuels 2017, 31, 14–36. [CrossRef] 11. Laredo, G.C.; Vega-Merino, P.M.; Trejo-Zarraga, F.; Castillo, J. Denitrogenation of middle distillates using adsorbent materials towards ULSD production. Fuel Process. Echnol. 2013, 106, 21–32. [CrossRef] 12. García-Gutiérrez, J.L.; Laredo, G.C.; García-Gutiérrez, P.; Jiménez-Cruz, F. Oxidative desulfurization of diesel using promising heterogeneous tungsten catalysts and hydrogen peroxide. Fuel 2014, 138, 118–125. [CrossRef] 13. Rana, M.S.; Al-Barood, A.; Brouresli, R.; Al-Hendi, A.W.; Mustafa, N. Effect of organic nitrogen compounds on deep hydrodesulfurization of middle distillate. Fuel Process. Technol. 2018, 177, 170–178. [CrossRef] 14. Liu, Y.; Liu, S.; Liu, S.; Liang, D.; Li, S.; Tang, Q.; Wang, X.; Miao, J.; Shi, Z.; Zheng, Z. Facile Synthesis of a Nanocrystalline Metal–Organic Framework Impregnated with a Phosphovanadomolybdate and Its Remarkable Catalytic Performance in Ultradeep Oxidative Desulfurization. ChemCatChem 2013, 5, 3086–3091. [CrossRef] 15. Julião, D.; Gomes, A.C.; Pillinger, M.; Cunha-Silva, L.; de Castro, B.; Gonçalves, I.S.; Balula, S.S. Desulfurization of model diesel by extraction/oxidation using a zinc-substituted polyoxometalate as catalyst under homogeneous and heterogeneous (MIL-101(Cr) encapsulated) conditions. Fuel Process. Technol. 2015, 131, 78–86. [CrossRef] 16. Julião, D.; Gomes, A.C.; Pillinger, M.; Valença, R.; Ribeiro, J.C.; de Castro, B.; Gonçalves, I.S.; Cunha Silva, L.; Balula, S.S. Zinc-Substituted Polyoxotungstate@amino-MIL-101(Al)—An Efficient Catalyst for the Sustainable Desulfurization of Model and Real Diesels. Eur. J. Inorg. Chem. 2016, 2016, 5114–5122. [CrossRef] 17. Granadeiro, C.M.; Nogueira, L.S.; Julião, D.; Mirante, F.; Ananias, D.; Balula, S.S.; Cunha-Silva, L. Influence of a porous MOF support on the catalytic performance of Eu-polyoxometalate based materials: Desulfurization of a model diesel. Catal. Sci. Technol. 2016, 6, 1515–1522. [CrossRef] 18. Xu, C.; Fang, R.; Luque, R.; Chen, L.; Li, Y. Functional metal–organic frameworks for catalytic applications. Coord. Chem. Rev. 2019, 388, 268–292. [CrossRef] 19. Shimizu, G.K.H.; Vaidhyanathan, R.; Taylor, J.M. Phosphonate and sulfonate metal organic frameworks. Chem. Soc. Rev. 2009, 38, 1430–1449. [CrossRef] 20. Firmino, A.D.G.; Figueira, F.; Tome, J.P.C.; Almeida Paz, F.A.; Rocha, J. Robust Metal-Organic Frameworks assembled from tetraphosphonic ligands and lanthanides. Coord. Chem. Rev. 2018, 355, 133–149. [CrossRef] 21. Pereira, C.F.; Figueira, F.; Mendes, R.F.; Rocha, J.; Hupp, J.T.; Farha, O.K.; Simoes, M.M.Q.; Tome, J.P.C.; Paz, F.A.A. Bifunctional Porphyrin-Based Nano-Metal-Organic Frameworks: Catalytic and Chemosensing Studies. Inorg. Chem. 2018, 57, 3855–3864. [CrossRef][PubMed] Catalysts 2020, 10, 731 15 of 15

22. Mendes, R.F.; Silva, P.; Antunes, M.M.; Valente, A.A.; Paz, F.A.A. Sustainable synthesis of a catalytic active one-dimensional lanthanide-organic coordination polymer. Chem. Commun. 2015, 51, 10807–10810. [CrossRef][PubMed] 23. Firmino, A.D.G.; Mendes, R.F.; Antunes, M.M.; Barbosa, P.C.; Vilela, S.M.F.; Valente, A.A.; Figueiredo, F.M.L.; Tome, J.P.C.; Paz, F.A.A. Robust Multifunctional Yttrium-Based Metal Organic Frameworks with Breathing Effect. Inorg. Chem. 2017, 56, 1193–1208. [CrossRef][PubMed] 24. Mendes, R.F.; Antunes, M.M.; Silva, P.; Barbosa, P.; Figueiredo, F.; Linden, A.; Rocha, J.; Valente, A.A.; Almeida Paz, F.A. A Lamellar Coordination Polymer with Remarkable Catalytic Activity. Chem. A Eur. J. 2016, 22, 13136–13146. [CrossRef] 25. Otsuki, S.; Nonaka, T.; Takashima, N.; Qian, W.; Ishihara, A.; Imai, T.; Kabe, T. Oxidative Desulfurization of Light Gas Oil and Vacuum Gas Oil by Oxidation and Solvent Extraction. Energy Fuels 2000, 14, 1232–1239. [CrossRef] 26. Ribeiro, S.; Barbosa, A.D.S.; Gomes, A.C.; Pillinger, M.; Gonçalves, I.S.; Cunha-Silva, L.; Balula, S.S. Catalytic oxidative desulfurization systems based on Keggin phosphotungstate and metal-organic framework MIL-101. Fuel Process. Technol. 2013, 116, 350–357. [CrossRef] 27. Xu, J.; Zhao, S.; Chen, W.; Wang, M.; Song, Y.-F. Highly Efficient Extraction and Oxidative Desulfurization System Using Na7H2LaW10O36 32 H2O in [bmim]BF4 at Room Temperature. Chem. A Eur. J. 2012, 18, · 4775–4781. [CrossRef] 28. Zhu, W.; Huang, W.; Li, H.; Zhang, M.; Jiang, W.; Chen, G.; Han, C. Polyoxometalate-based ionic liquids as catalysts for deep desulfurization of fuels. Fuel Process. Technol. 2011, 92, 1842–1848. [CrossRef] 29. Ribeiro, S.O.; Nogueira, L.S.; Gago, S.; Almeida, P.L.; Corvo, M.C.; Castro, B.D.; Granadeiro, C.M.; Balula, S.S. Desulfurization process conciliating heterogeneous oxidation and liquid extraction: Organic solvent or centrifugation/water? Appl. Catal. A Gen. 2017, 542, 359–367. [CrossRef] 30. Julião, D.; Gomes, A.C.; Pillinger, M.; Valença, R.; Ribeiro, J.C.; Gonçalves, I.S.; Balula, S.S. Desulfurization of liquid fuels by extraction and sulfoxidation using H2O2 and [CpMo(CO)3R] as catalysts. Appl. Catal. B Environ. 2018, 230, 177–183. [CrossRef] 31. Jiao, T.; Zhuang, X.; He, H.; Zhao, L.; Li, C.; Chen, H.; Zhang, S. An ionic liquid extraction process for the separation of indole from wash oil. Green Chem. 2015, 17, 3783–3790. [CrossRef] 32. Liu, D.; Gui, J.; Sun, Z. Adsorption structures of heterocyclic nitrogen compounds over Cu(I)Y zeolite: A first principle study on mechanism of the denitrogenation and the effect of nitrogen compounds on adsorptive desulfurization. J. Mol. Catal. A Chem. 2008, 291, 17–21. [CrossRef] 33. Zhou, Z.Q.; Li, W.S.; Liu, J. Removal of nitrogen compounds from fuel oils using imidazolium-based ionic liquids. Pet. Sci. Technol. 2017, 35, 45–50. [CrossRef] 34. Xie, L.-L.; Favre-Reguillon, A.; Wang, X.-X.; Fu, X.; Pellet-Rostaing, S.; Toussaint, G.; Geantet, C.; Vrinat, M.; Lemaire, M. Selective extraction of neutral nitrogen compounds found in diesel feed by 1-butyl-3-methyl-imidazolium chloride. Green Chem. 2008, 10, 524–531. [CrossRef] 35. Hansmeier, A.R.; Meindersma, G.W.; de Haan, A.B. Desulfurization and denitrogenation of gasoline and diesel fuels by means of ionic liquids. Green Chem. 2011, 13, 1907–1913. [CrossRef] 36. Mendes, R.F.; Barbosa, P.; Domingues, E.M.; Silva, P.; Figueiredo, F.; Paz, F.A.A. Enhanced proton conductivity in a layered coordination polymer. Chem. Sci. 2020. In Press. [CrossRef] 37. Silva, P.; Mendes, R.F.; Fernandes, C.; Gomes, A.C.; Ananias, D.; Remião, F.; Borges, F.; Valente, A.A.; Paz, F.A.A. Multifunctionality and cytotoxicity of a layered coordination polymer. Dalton Trans. 2020, 49, 3989–3998. [CrossRef]

© 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/).