The Heterogeneous Aminohydroxylation Reaction

catalysts Article

The Heterogeneous Aminohydroxylation Reaction Using Hydrotalcite-Like Catalysts Containing Osmium Mohamed I. Fadlalla 1,2 , Glenn E. M. Maguire 1 and Holger B. Friedrich 1, * 1 2

*

School of Chemistry and Physics, University of KwaZulu-Natal, Durban 4041, South Africa; [email protected] (M.I.F.); [email protected] (G.E.M.M.) Catalysis Institute, Department of Chemical Engineering, University of Cape Town, Rondebosch 7701, South Africa Correspondence: [email protected]; Tel.: +27-(0)31-260-3107

Received: 30 September 2018; Accepted: 30 October 2018; Published: 16 November 2018

Abstract: The aminohydroxylation reaction of olefins is a key organic transformation reaction, typically carried out homogeneously with toxic and expensive osmium (Os) catalysts. Therefore, heterogenisation of this reaction can unlock its industrial potential by allowing reusability of the catalyst. Os–Zn–Al hydrotalcite-like compounds (HTlcs), as potential heterogeneous aminohydroxylation catalysts, were synthesised by the co-precipitation method and characterised by several techniques. Reaction parameters (i.e., solvent system, reaction temperature, and catalyst structure) were optimized with cyclohexene, styrene, and hexene as substrates. The different classes of olefins (aliphatic, aromatic, and functionalised) that were tested gave >99% conversion and high selectivity (>97%) to the corresponding β-amino alcohol. The catalyst HTlc structure had a significant effect on the reaction time and yield of the β-amino alcohols. Under the same testing conditions, a heat treated catalyst (non-HTlc) showed a shorter reaction time, but drop in the yield of β-amino alcohols and rise in diol formation was observed. Leaching tests showed that 2.9% and 3.4% of Os (inactive) leached from the catalyst to the reaction solution when MeCN/water (1:1 v/v) and t-BuOH/water (1:1 v/v), respectively, were used as the solvent system. Recycling studies showed that the catalyst can be reused at least thrice, with no significant difference in the yield of the β-amino-alcohol. Keywords: aminohydroxylation; heterogeneous catalysis; hydrotalcite-like compounds; leaching and recyclability study

1. Introduction Olefins are considered one of the most important starting materials in organic synthesis. They are cheap and widely available, and the double bond allows 1,2-functionalisation via face selective oxidation [1,2]. For example, olefins are utilised as starting materials to produce diols and vicinal amino alcohols during the dihydroxylation (DH) and aminohydroxylation (AH) reaction, respectively. Both reactions (i.e., DH and AH) are catalysed by OsO4 and are considered to be highly selective and reliable organic transformations, because OsO4 reacts with a wide spectrum of olefins and does not react or reacts very slowly with other organic functional groups present [3]. The AH reaction has been studied extensively by Sharpless et al. [4,5], and more recently, considerable interest has focused on the AH reaction, as β-amino alcohols are an important functionality in chiral ligands and biologically active compounds [6]. β-Amino alcohols are classified into three general groups, including (i) naturally occurring compounds, (ii) synthetic pharmacologically active molecules, and (iii) chiral ligands [7–9]. The homogeneous asymmetric aminohydroxylation reaction uses both osmium and, usually, a cinchona alkaloid ligand, both of which are relatively expensive components when trying to scale up the reaction. Furthermore, osmium is a toxic metal, hence there have been investigations on the Catalysts 2018, 8, 547; doi:10.3390/catal8110547

www.mdpi.com/journal/catalysts

Catalysts 2018, 8, 547

2 of 13

aminohydroxylation reaction using non-toxic or less toxic metals such as iron and copper [10,11]. There have been several attempts at heterogenising the asymmetric aminohydroxylation reaction (using osmium) to allow the re-use of these components. One approach was by supporting the 1,4-bis(9-O-quininyl)phthalazine ((QN)2 PHAL) ligand on silica gel [12]. The results obtained with this method showed percentage yields of β-amino alcohols ranging from 30 to 81% (based on the nature of the olefin) and enantiomeric excess (ee) values ranging from 88 to >99%. However, this system resulted in osmium leaching into the product and the addition of osmium was required to restore the original activity [12]. A different approach was to support the (QN)2 PHAL on a polymer, which resulted in percentage conversions to β-amino alcohols ranging from 73 to 98% and ee ranging from 83 to 87%. However, upon recycling of the catalyst, the percentage conversion reduced to 58% and the ee value to 81% [13]. Unlike the silica supported catalyst, the activity was not restored by further addition of osmium. The latest approach to heterogenising this reaction involved supporting OsO4 on a layered double hydroxide (LDH). This resulted in yields of β-amino alcohols ranging from 45 to 55% and ee values ranging from 40 to 78%. However, high levels of osmium leaching were observed. The authors suggested that this could be the result of the high polarizing effect of chloramine-T that was used as the nitrogen source [14]. Previous studies, apart from also forming diol by-products, have two facts in common, osmium leaching and the reasonable to excellent enantiomeric excess attained. Furthermore, the approaches that heterogenise the ligand do not immobilise the osmium [12,13]. Therefore, the problem of heterogenising the osmium remains, which is the focus of this study. Hydrotalcite (HT) and hydrotalcite-like compounds (HTlcs) are examples of anionic clays and can be defined as layered, positively charged brucite-like structures in which the net charge is balanced by anions in the interlayer [15,16]. HT consists of hydroxyl-carbonates of aluminum and magnesium [12,17]. Since their discovery in 1842, hydrotalcites are also referred to as layered double hydroxides [12,16–18]. HTlcs are compounds that consist of metals other than aluminum and magnesium, but exhibit the same structure as a hydrotalcite [16]. HTlcs can be tailored for specific catalytic application by changing the elements and the ratio in which they are present. Furthermore, the active cations can be isomorphically substituted with Mg2+ or Al3+ in the octahedral site. Previous research on the use of Os–HTlc catalysts in the heterogenization of the dihydroxylation reaction [19] showed that they have activity, selectivity, and catalyst stability. These results inspired us to utilize the Os–HTlc system for the current study of the heterogenization of the aminohydroxylation reaction. This study thus focused on heterogenising the aminohydroxylation reaction by incorporating osmium as part of a hydrotalcite-like structure. The effects of different conditions (temperature and solvent system), the catalyst crystallinity, and structure were examined. 2. Results 2.1. Catalyst Characterization The Os–HTlc catalyst was prepared via the co-precipitation method with a target ratio of 0.3 Os/Al, based on previously reported studies [19,20]. The catalyst characterisation focused on determining the metal content, phase composition, surface area, and metal distribution in the catalysts. The three metals present (Os, Zn, and Al) in the prepared materials were quantitatively determined by means of inductively coupled plasma optical emission spectroscopy (ICP-OES) (Table 1). In catalyst A, the ratio of osmium to aluminum was close to the target of 0.3. This ratio is reported to give a HTlc structure [19]. Furthermore, the ratio of 0.3 Os/Al gave a catalyst with good activity in the dihydroxylation reaction, as previously reported [20]. However, the ratio of zinc to aluminum was slightly higher than the target value of 3. After the heat treatment at 300 ◦ C for two hours of catalyst A, to give catalyst B, the osmium content decreased by 42%, which was likely the result of sublimation. As a consequence the HTlc structure collapsed, which resulted in a mixed metal oxide.

Catalysts 2018, 8, x FOR PEER REVIEW

3 of 12


3 of 13

catalyst A, to give catalyst B, the osmium content decreased by 42%, which was likely the result of sublimation. As a consequence the HTlc structure collapsed, which resulted in a mixed metal oxide. The The powder powder XRD XRD pattern pattern for for the the crystalline crystalline HTlc HTlc (Figure (Figure 1a) 1a) isis characterised characterised by by several several high high intensity peaks at low 2θ values. As the values for 2θ increase, the intensity of the peaks intensity peaks at low 2θ values. As the values for 2θ increase, the intensity of the peaks decreases, decreases, in in the the plane plane order order of of 003 003 to to 110. 110. The The equal equal spacing spacing between between planes planes 003 003 and and 006, 006, and and 006 006 and and 012, 012, indicate the stacking order of the HTlc’s. HT and HTlc’s have a hexagonal plane, which is the indicate the stacking order of the HTlc’s. HT and HTlc’s have a hexagonal plane, which is the 110 110 plane, shown by a doublet peak at high 2θ value and low intensity. The presence of a hydroxide plane, shown by a doublet peak at high 2θ value and low intensity. The presence of a hydroxide in in material is indicated a peak a d-spacing Å. The sharpness of the peaks is thethe material is indicated by aby peak with awith d-spacing of 11.1 of Å . 11.1 The sharpness of the peaks is indicative indicative of the crystallinity of the catalyst. The XRD pattern of catalyst B (Figure 1b) showed that the of the crystallinity of the catalyst. The XRD pattern of catalyst B (Figure 1b) showed that the compound did not maintain the HTlc structure after heating. This is indicated by the absence of all the compound did not maintain the HTlc structure after heating. This is indicated by the absence of all characteristic features in the expected of the The XRD pattern (Figure 1b) shows that the characteristic features inpattern the pattern expected ofHTlc. the HTlc. The XRD pattern (Figure 1b) shows there is no crystalline HTlc phase present. According to literature, osmium can migrate to the surface that there is no crystalline HTlc phase present. According to literature, osmium can migrate to the upon heating, waswhich evidentwas in HRTEM Furthermore, of the hydrotalcite-like surface upon which heating, evident[20]. in HRTEM [20].calcination Furthermore, calcination of the ◦ C can result in the presence of both the HT phase and the metal oxide phase [21]. compound at 300 hydrotalcite-like compound at 300 °C can result in the presence of both the HT phase and the metal Possibly, in part, a result in of part, the low the crystallinity heat-treated of catalyst, this was not observed oxide phase [21].as Possibly, as acrystallinity result of theoflow the heat-treated catalyst, this for B. wascatalyst not observed for catalyst B.

Figure 1. XRD pattern of Os–hydrotalcite-like compound (HTlc) (a) and Os–HTlc treated at 300 ◦ C Figure 1. XRD pattern of Os–hydrotalcite-like compound (HTlc) (a) and Os–HTlc treated at 300 °C (b) (b) obtained with Cu source (λ = 1.5406 Å). obtained with Cu source (λ = 1.5406 Å). Table andand physical properties of temperature (un)treated Os-HTlc prepared co-precipitation Table1.1.Chemical Chemical physical properties of temperature (un)treated Os-HTlcbyprepared by cofor the aminohydroxylation reaction of olefins. precipitation for the aminohydroxylation reaction of olefins. Catalyst Catalyst Catalyst Catalyst A A Catalyst B Catalyst B

BET-SURFACE Crystallite a c BET-SURFACE Area/m2 Crystallite Size/Å a Parameter/Å c Parameter/Å −1 Size/Å Parameter/Å Parameter/Å g Area/m2 g−1 196 196 -

-

3.06 -

3.06 -

23.09 -

23.09 -

87 150

87 150

ICP-OES ICP-OES ofof Os/Zn/Al Os/Zn/Al 0.29/3.5/1 0.29/3.5/1 0.17/3.56/1

0.17/3.56/1

The Thesurface surfacemorphology morphologyof ofcatalyst catalystA Aand andthe the distribution distributionof ofthe thethree threemetals metalsin inthe the catalyst catalystwas was investigated investigated by by means means of of scanning scanning electron microscopy (SEM) and electron dispersion spectroscopy, respectively. respectively. The The SEM SEM images images of of catalyst catalyst A A (Figure (Figure 2a) 2a) show show that that the the catalyst catalyst isis crystalline, crystalline, which which correlates with the XRD results mentioned previously. As the sample was ground and selected correlates with the XRD results mentioned previously. As the sample was ground and selected randomly randomly for for the the EDS EDS analysis, analysis, one one can can assume assume that that the the sample sample is homogeneous. More importantly, EDS images show that osmium is homogeneously distributed EDS images show that osmium is homogeneously distributed in in the the HTlc HTlc (Figure (Figure 2b), 2b), with with no no agglomeration agglomerationdetected, detected,which whichfurther furtherconfirms confirmsthat thatosmium osmiumisispart part of of the the catalyst catalyst structure structureand and not not just just supported supportedon onthe theHTlc. HTlc.

Catalysts 2018, 8, x FOR PEER REVIEW


Catalysts 2018, 8, x8,FOR PEER REVIEW Catalysts 2018, x FOR PEER REVIEW

4 of 12

4 of 13 4 of 124 of 12

Figure 2. Surface morphology of Os–HTlc (a) and EDX mapping of Os in the Os–HTlc catalyst (b).

2.2. Catalytic Testing Surface morphology (a) and EDX of Os in the Os–HTlc catalyst (b). in the The Figure next 2.step focused onof Os–HTlc investigating themapping Os–HTlc catalyst performance Figure 2. Surface morphology of Os–HTlc Os–HTlc mapping of Os in solvent, the catalyst (b). (b). aminohydroxylation reaction (Scheme 1) as(a) a and function of reaction temperature, catalyst Figure 2. Surface morphology of (a) andEDX EDX mapping of Os in Os–HTlc the Os–HTlc catalyst 2.2. Catalytic Testing structure, and finally the suitability of the catalyst in the aminohydroxyation of different olefins at 2.2. Catalytic Testing Theconditions. next step focused on investigating the Os–HTlc catalyst performance in the 2.2.optimal Catalytic Testing the

The next step focused on investigating catalyst performance in the aminohydroxylation aminohydroxylation reaction (Scheme 1)the asOs–HTlc a function of reaction solvent, temperature, catalyst The next step focused on investigating the Os–HTlc catalyst performance in the reaction (Scheme 1) asthe a function of reaction solvent,intemperature, catalyst structure, and finally structure, and finally suitability of the catalyst the aminohydroxyation of different olefinsthe at aminohydroxylation reaction (Scheme 1) as a function of reaction solvent, temperature, catalyst suitability ofconditions. the catalyst in the aminohydroxyation of different olefins at the optimal conditions. the optimal

structure, and finally the suitability of the catalyst in the aminohydroxyation of different olefins at the optimal conditions.

Scheme 1. Heterogeneous aminohydroxylation reaction over Os–hydrotalcite-like compound (HTlc) at 25 °C with chloramine-T as nitrogen source (Ts = tosyl group). Scheme compound (HTlc) at Scheme 1. 1. Heterogeneous Heterogeneousaminohydroxylation aminohydroxylationreaction reactionover overOs–hydrotalcite-like Os–hydrotalcite-like compound (HTlc) ◦ C with chloramine-T as nitrogen source (Ts = tosyl group). 25 at 25 with chloramine-T as nitrogen 2.2.1. Effect of°C Solvent on Reaction Time source (Ts = tosyl group).

2.2.1. Effect Solvent on Reaction The solvents that were chosenTime for this study (i.e., toluene, MeCN/water (1:1 v/v), and tScheme 1. of Heterogeneous aminohydroxylation reaction over Os–hydrotalcite-like compound (HTlc) 2.2.1. Effect of Solvent on Reaction Time at 25 °C with chloramine-T as nitrogen source (Ts = tosyl group). BuOH/water (1:1 v/v)) display a wide range of polarities [22]. The results obtained (Figure 3) The solvents that were chosen for this study (i.e., toluene, MeCN/water (1:1 v/v), and t-BuOH/watershow The solvents that were chosen for this study (i.e., toluene, MeCN/water (1:1 v/v), and tthat the timeaincreases following MeCN/water (1:1 v/v) t-BuOH/water (1:1 v/v) (1:1 reaction v/v)) display wide rangeinofthe polarities [22].order: The results obtained (Figure 3) ≈show that the reaction BuOH/water (1:1 v/v)) display a wide range of polarities [22]. The results obtained (Figure 3) show 2.2.1. Effect of Solvent on Reaction Time < toluene, and theindielectric constants for these solvents in the same order. time increases the following order: MeCN/water (1:1 v/v)decrease ≈ t-BuOH/water (1:1 v/v) < toluene, and the that the reaction time increases in the following order: MeCN/water (1:1 v/v) ≈ t-BuOH/water (1:1 v/v) dielectric constants for these solvents decrease in the same order. < The toluene, and thethat dielectric decrease in the same order. solvents wereconstants chosen for forthese this solvents study (i.e., toluene, MeCN/water (1:1 v/v), and t-

BuOH/water (1:1 v/v)) display a wide range of polarities [22]. The results obtained (Figure 3) show that the reaction time increases in the following order: MeCN/water (1:1 v/v) ≈ t-BuOH/water (1:1 v/v) < toluene, and the dielectric constants for these solvents decrease in the same order.

Figure 3. Effect of the solvent system on the reaction time for 99.99% conversion of the olefin determined

Figure 3. Effect of the solventsystem systemon onthe the reaction reaction time of the olefin Figure 3. Effect of the solvent time for for99.99% 99.99%conversion conversion of the olefin by gas chromatograph (GC), using chloramine-T as the nitrogen source and catalyst A at catalyst 60 ◦ C. A at determined by gas chromatograph (GC), using chloramine-T as the nitrogen source and determined by gas chromatograph (GC), using chloramine-T as the nitrogen source and catalyst A at 60 °C.60 °C.

Figure 3. Effect of the solvent system on the reaction time for 99.99% conversion of the olefin determined by gas chromatograph (GC), using chloramine-T as the nitrogen source and catalyst A at 60 °C.

Catalysts2018, 2018, 8, 8, x547 Catalysts FOR PEER REVIEW

13 5 5ofof12

2.2.2. Effect of Temperature on Reaction Time 2.2.2. Effect of Temperature on Reaction Time This investigation was carried out at temperatures of 25 °C and 60 °C using the three solvent This investigation was carried out at temperatures of 25 ◦ C and 60 ◦ C using the three solvent systems and styrene as model olefin (Figure 4). Temperature did not have a significant effect on either systems and styrene as model olefin (Figure 4). Temperature did not have a significant effect on either reaction time or product distribution when MeCN/water and t-BuOH/water (1:1 v/v) were used. Only reaction time or product distribution when MeCN/water and t-BuOH/water (1:1 v/v) were used. Only trace amounts of the diol were detected when these two solvent systems were employed with catalyst trace amounts of the diol were detected when these two solvent systems were employed with catalyst A. In the case of toluene, the reaction time decreased with an increase in reaction temperature up to A. In the case of toluene, the reaction time decreased with an increase in reaction temperature up to 60 °C. This is possibly the result of the fact that chloramine-T is completely soluble in MeCN/water 60 ◦ C. This is possibly the result of the fact that chloramine-T is completely soluble in MeCN/water and t-BuOH/water (1:1 v/v), but is only partially soluble in toluene. Thus, an increase in the reaction and t-BuOH/water (1:1 v/v), but is only partially soluble in toluene. Thus, an increase in the reaction temperature increases the solubility and hence decreases the reaction time. temperature increases the solubility and hence decreases the reaction time. 25 ˚C

60 ˚C

60

Time /h

50 40 30 20 10 0 Toluene

MeCN/H2O

t-BuOH/H2O

Figure 4. Effect of temperature on reaction time (complete depletion of starting material), using styrene, Figure 4. Effect of temperature on reaction time (complete depletion of starting material), using chloramine-T as the nitrogen source, and catalyst A. styrene, chloramine-T as the nitrogen source, and catalyst A.

2.2.3. Effect of Catalyst Structure on Reaction Time 2.2.3. Effect of Catalyst Structure on Reaction Time The comparison between the HTlc (catalyst A) and the mixed metal oxide (catalyst B) in the The comparison reaction between should the HTlc A) and thetomixed metal the oxide (catalyst the aminohydroxylation aid(catalyst in understanding what extent structure ofB) theinHTlc aminohydroxylation reaction should aid in understanding to what structure of structure the HTlc is is necessary for catalytic activity in this reaction, as literature hasextent shownthe that the HTlc is necessary for catalytic activity in this reaction, as literature has shown that the HTlc structure is important for a number of reactions [19,20]. Thus, catalyst B, which had lost the HTlc structure, was important forto a number of reactions catalyst B, whichactivity. had lostThese the HTlc structure,were was also studied investigate the effect[19,20]. of heat Thus, treatment on catalytic experiments also studied to investigate the effect heat treatmentason activity. These experiments carried out utilizing cyclohexene and of t-butylcrotonate thecatalytic model olefins with MeCN/water (1:1were v/v) carried out utilizing cyclohexene and t-butylcrotonate as the model olefins with MeCN/water (1:1 as the solvent system. The reaction time for cyclohexene was 6 h and 3 h, over catalysts A and B, v/v) as the solvent The reaction time for cyclohexene was h and 3 h,both overcatalysts. catalystsHowever, A and B, respectively, whilesystem. for t-butylcrotonate, the reaction time was ca.622 h over respectively, for t-butylcrotonate, the reaction time was ca. 22increased h over both catalysts. the selectivitywhile to β-amino alcohols dropped and selectivity to diols over catalyst However, B. the selectivity to β-amino alcohols dropped and selectivity to diols increased over catalyst B. 2.2.4. Screening of Different Classes of Olefins 2.2.4. Screening of Different Classes of Olefins The olefins utilized in this study allow the investigation of alkene nature and steric effects on the The olefins in (Scheme this study allow 2). theThe investigation of alkene nature and steric effects on reaction time andutilized the yield 1, Table substrates can be divided into aliphatic, aromatic, ◦ the time and the yield (Scheme 1, were Tablecarried 2). Theout substrates andreaction functionalised olefins. All the reactions at 25 C. can be divided into aliphatic, aromatic, and functionalised olefins.toAll reactions carried outthan at 25for °C.the functionalised ones, The reaction time was observed be the shorter for thewere aliphatic olefins reaction time observed to be shorter for the aliphatic olefins than and for the functionalised withThe hexene having thewas shortest reaction time and cis-stilbene, methylcinnamate, dimethylfumarate ones, hexene reaction having times. the shortest reaction and cis-stilbene, methylcinnamate, and havingwith the longest Furthermore, the time two solvent systems show comparable reaction dimethylfumarate having the longest reaction times. Furthermore, the two solvent systems show times for the same substrate. Similar observations were reported in the homogeneous reaction [13,14]. comparable times for the same substrate. observations were reported in the Selectivity toreaction the aminohydroxylation products exceeds Similar 99% in all cases. homogeneous reaction [13,14]. Selectivity to the aminohydroxylation products exceeds 99% in all cases.

Catalysts x FOR 6 6ofof1212 Catalysts2018, 2018,8, FORPEER PEERREVIEW REVIEW Catalysts 2018, 8,8,x xFOR PEER REVIEW 6 of 12 Catalysts 2018, 8, x FOR PEER REVIEW 6 of 12 Catalysts 2018, 8, 547 6 of 13 Catalysts 2018, 8, x FOR PEER REVIEW 6 of 12 Catalysts 2018, 8, x FOR PEER REVIEW 6 of 12 Table time and Catalysts 2018,2.2. 8,Reaction xReaction FOR PEER REVIEW 6 of 12 Table time andhigh highperformance performanceliquid liquidchromatograph chromatograph(HPLC) (HPLC)yield yieldpercentage percentagefor forthe the

Table 2. Reaction time and high performance liquid chromatograph (HPLC) yield percentage for the Table 2. Reaction timeinand high performance liquid chromatograph (HPLC) yield percentage for the different olefins aminohydroxylation reaction catalysed by different olefinsused used inthe the aminohydroxylation reaction catalysed(HPLC) bythe theHTlc HTlccatalyst catalystcontaining containing Table 2. Reaction Reaction timeinand and high performance liquid liquid chromatograph (HPLC) yield percentage for the the different olefins used the aminohydroxylation reaction catalysed by the HTlc catalyst containing Table 2. time high performance chromatograph yield percentage for Table different 2. Reaction time and high performance liquid chromatograph (HPLC) yield percentage for the olefins used in the aminohydroxylation reaction catalysed by the HTlc catalyst containing osmium at 25 °C. Table 2. Reaction time and high performance liquid chromatograph (HPLC) yield percentage for the osmium at 25 °C. different at olefins used in in the the aminohydroxylation aminohydroxylation reaction reaction catalysed catalysed by by the the HTlc HTlc catalyst catalyst containing containing osmium 25 °C.used different olefins different olefins used the in aminohydroxylation reaction catalysed by theby HTlc containing osmium at 25 °C.in different olefins used the aminohydroxylation reaction catalysed the catalyst HTlc catalyst containing osmium at 25 °C. Reaction Time/h in MeCN/H 2O % Yield of β-Amino ◦ osmium °C. ReactionTime/h Time/hininMeCN/H MeCN/H 2O Yieldofofβ-Amino β-AminoAlcohol, Alcohol,in osmium at 25 at C.25 Entry Reaction 2O %%MeCN/H Yield Alcohol, inin osmium at 25 Olefin °C. Entry Olefin 2O) 2O (t-BuOH/H 2O) in Reaction(t-BuOH/H Time/h in MeCN/H 2O % Yield of β-Amino Alcohol,

Entry Entry Entry Entry Entry Entry 11 1 1 1 1 11 22 2 2 2 2 22 33 3 3 3 33 3 44 4

4 44 4 4

55

Olefin Olefin Olefin Olefin Olefin Olefin

(t-BuOH/H 2O) Reaction(t-BuOH/H Time/h in 2MeCN/H 2O O) Reaction(t-BuOH/H Time/h in 2MeCN/H 2O O) Reaction Time/h Reaction Time/h in MeCN/H 2O (t-BuOH/H 2O)in 2O) MeCN/H(t-BuOH/H (t-BuOH/H 7 7(8)(8) 2O) 2 O) 2 O (t-BuOH/H 7 (8) 7 (8) 7 (8) 7 (8) (8) 7 7(8) 9.5 9.5(10) (10) 9.5 (10) 9.5 (10) 9.5 (10) 9.5(10) (10) 9.5 9.5 (10) 7 7(6)(6) 7 (6) 7 (6) (6) 7 77(6) (6) 7 (6) 2424(24) (24) 24 (24)

(24) 2424 24(24) (24) 24 (24) 24 (24)

2323(24) (24)

MeCN/H 2O (t-BuOH/H 2O) % MeCN/H Yield of β-Amino Alcohol, 2O (t-BuOH/H 2O) in

% MeCN/H Yield of β-Amino Alcohol, 2O (t-BuOH/H 2O) in % Yield of β-Amino Alcohol, % Yield of β-Amino Alcohol, MeCN/H 2O (t-BuOH/H 2O) in MeCN/H2O (t-BuOH/H2O) in MeCN/H MeCN/H 2O 2O) 99(t-BuOH/H (99) 2 O (t-BuOH/H 2 O) (99) 9999(99) 99 (99) 99 (99) 99 (99) 99 (99) 99 (99)

9999(99) (99) 99 (99) 99 (99) 99 (99) 99 (99) 99 (99) 99 (99) 9999(99) (99) 99 (99) 99 (99) 99 (99) 99 (99) 99 (99) 99 (99) 9999(99) (99) 99 (99)

99 (99) 99 (99) 99 (99) 99 (99) 99 (99)

9999(99) (99)

5 55

23(24) (24) 2323 (24)

6 6 66

24(24) (24) 2424(24)

(99) 99 (99) 9999(99)

7 77 7

2222(23) 22(23) (23)

9999(99) 99 (99) (99)

5 5 5

6 6 6 6

7 7 7 7

23 (24) 23 (24) 23 (24)

24 (24)

24 (24) 24 (24) 24 (24) 24 (24)

22 (23) 22 (23) 22 (23) 22 (23) 22 (23)

99 (99) 99 (99) 99 (99)

99 (99) 99 (99) 99 (99)

99 (99)

99 (99) 99 (99) 99 (99) 99 (99)

99 (99) 99 (99) 99 (99) 99 (99) 99 (99)

2.3. 2.3. Leaching Test Test 2.3.Leaching Leaching Test 2.3. Leaching Test 2.3. Leaching Test 2.3.the Leaching Testto be to For the heterogeneous, the not into the For reaction heterogeneous, the active species must must not leach into the reaction 2.3. Leaching Test For thereaction reactionstrictly tobe bestrictly strictly heterogeneous, theactive activespecies species must notleach leach into thereaction reaction 2.3. Leaching Test For the reaction to be strictly heterogeneous, the active species must not leach into the reaction For the reaction to be strictly heterogeneous, the active species must not leach into the reaction mixture, or, if they leach, their homogeneous form must not be catalytically active [23]. To determine mixture, or, if they leach, their homogeneous form must not be catalytically active [23]. To determine if For the the reaction to be betheir strictly heterogeneous, the active species must not not leach intoTo the reaction mixture, or,if ifthey theyleach, leach, their homogeneous form must notbe becatalytically catalytically active [23]. To determine mixture, or, homogeneous form must not active [23]. determine For reaction to strictly heterogeneous, the active species must leach into the reaction mixture, or, if they leach, their homogeneous form must not be catalytically active [23]. To determine For the reaction to be strictly heterogeneous, the active species must not leach into the reaction if the reaction is heterogeneous, catalyst A was removed after approximately 20% of the cyclohexene the reaction is heterogeneous, catalyst A was removed after approximately 20% of the cyclohexene mixture, or, if if they they leach, their their homogeneous homogeneous form must not not beapproximately catalytically active [23]. To determine ifthe thereaction reaction heterogeneous, catalystAAwas was removed after approximately 20%of ofthe the cyclohexene if isisheterogeneous, heterogeneous, catalyst removed after 20% cyclohexene mixture, or, leach, form must be catalytically active [23]. To determine ◦ C. if the reaction is catalyst A was removed after 20% of the cyclohexene mixture, or, if they their homogeneous form must not beapproximately catalytically [23]. had reacted (after 2leach, h) at 25 °C. The reaction mixture (without the catalyst) was allowed todetermine stir had reacted (after 2(after h)heterogeneous, at 25 The reaction mixture (without the catalyst) wasactive allowed toTo stir for if the reaction is catalyst A was removed after approximately 20% of the cyclohexene had reacted 2 h) at 25 °C. The reaction mixture (without the catalyst) was allowed stirfor foraaa had reacted (after 2 h) h) at at 25 25 °C. °C. The The reaction mixture (without the catalyst) catalyst) was was allowed totostir stir for if thereacted reaction is heterogeneous, catalyst A was removed after approximately 20%allowed of the cyclohexene had (after 2 reaction mixture (without the to for a if the reaction is heterogeneous, catalyst A was removed after approximately 20% of the cyclohexene further 6 h. ICP analysis of the solution showed that osmium does leach, a small degree, but not a further 6 h. ICP analysis of the solution showed that osmium does leach, to small degree, but had reacted (after h) at atof 25 °C.solution The reaction reaction mixture (without the leach, catalyst) was allowed tobut stir for further ICPanalysis analysis ofthe the solution showed thatosmium osmium does leach,totoawas asmall small degree, butnot notin further 66h. h.h.ICP ICP showed that does degree, in had reacted (after 22 h) 25 °C. The mixture (without the catalyst) allowed to stir for aain further 6 analysis of the solution showed that osmium does leach, to a small degree, but not in had reacted (after 2 h) at 25 °C. The reaction mixture (without the catalyst) was allowed to stir for a an active form, as gas chromatograph (GC) analysis of the solution showed no further starting not infurther an active asasgas chromatograph analysis of the solution showed no further starting h.form, ICP analysis analysis ofchromatograph the solution solution(GC) showed that osmium osmium does leach, to small degree, but not in in anactive active form, gas (GC) analysis the solution showed no further starting an as gas gas chromatograph (GC) analysis ofofthe the solution showed no further starting further 66 h. ICP of the showed that does leach, to aa small degree, but not an active form, as chromatograph (GC) analysis of solution showed no further starting further 6 h. ICP analysis of the solution showed that osmium does leach, to a small degree, but not in material depletion after removal the The of leached from catalyst material depletion after of theofcatalyst. Theanalysis amounts osmium leached fromno catalyst A starting in AA in an active form, asremoval gas chromatograph (GC) ofofthe the solution showed further material depletion afterchromatograph removal thecatalyst. catalyst. Theamounts amounts ofosmium osmium leached from catalyst material depletion after removal ofofthe the catalyst. The amounts of osmium leached from catalyst A in in an active form, as gas (GC) analysis of solution showed no further starting material depletion after removal of catalyst. The amounts of osmium leached from catalyst A an active form, as gas chromatograph (GC) analysis of the solution showed no further starting MeCN/water and t-BuOH/water was similar at 2.9% and 3.4%, respectively. This effect could MeCN/water and t-BuOH/water was similar at 2.9% and 3.4%, respectively. This effect could be the material depletion after removal of ofwas the catalyst. The amounts ofrespectively. osmium leached leached from catalyst Athe in MeCN/water andt-BuOH/water t-BuOH/water similar at2.9% 2.9%amounts and3.4%, 3.4%, respectively. Thiseffect effectcatalyst couldbe be the MeCN/water and was similar at and This could be the material depletion after removal the catalyst. The of osmium from A in and t-BuOH/water was similar at 2.9% and 3.4%, This effect could be material depletion after removal of the catalyst. The amounts ofrespectively. osmium leached from catalyst Athe in result the strong polarizing ofof chloramine-T [14]. resultMeCN/water of theof strong polarizing effect effect ofeffect chloramine-T [14]. MeCN/water and t-BuOH/water was similar at 2.9% and 3.4%, respectively. This effect could be the result of the strong polarizing chloramine-T [14]. result of of the the strong strong polarizing effect effect ofsimilar chloramine-T [14]. MeCN/water and t-BuOH/water wasof at 2.9% [14]. and 3.4%, respectively. This effect could be the result polarizing chloramine-T MeCN/water and t-BuOH/water wasthat similar atpreviously 2.9% and 3.4%, respectively. effect could the An important result this work over that reported [12–14,24] isisthat the lowest An important result of thisof work over previously reported [12–14,24] is that This the lowest level ofbelevel result of the strong polarizing effect of chloramine-T [14]. An important result of this work over that previously reported [12–14,24] that the lowest level An important result of this this workofover over that previously previously reported [12–14,24] is that that the lowest level resultAn of important the strong result polarizing effect chloramine-T [14]. reported of work that [12–14,24] is the lowest level result of the strong polarizing effect of chloramine-T [14]. of inactive osmium leaching to the reaction mixture was observed. Previous attempts atat inactive osmium leaching to the reaction mixture was observed. Previous attempts at heterogenisation An important important result of this this work work overreaction that previously previously reported [12–14,24]Previous isPrevious that the the attempts lowest level inactive osmium leaching the reaction mixture was observed. observed. attempts ofof inactive inactive osmium leaching toto the the mixture was at An result of over that reported [12–14,24] is that lowest level of osmium leaching to reaction mixture was observed. Previous attempts at An important result of this work over that previously reported [12–14,24] is that the lowest level ofofthe aminohydroxylation reaction had leaching ranging from 2020toto50% of theheterogenisation aminohydroxylation reaction had leaching ranging from to 50% [12]. of inactive osmium osmium leaching toshown the reaction reaction mixture was 20observed. observed. Previous attempts at heterogenisation the aminohydroxylation reaction hadshown shown leaching ranging from 50% heterogenisation of the aminohydroxylation reaction had shown leaching ranging from 20 to 50% of inactive leaching to the mixture was Previous attempts at heterogenisation of the aminohydroxylation reaction had shown leaching ranging from 20 to 50% of inactive osmium leaching to the reaction mixture was observed. Previous attempts at [12]. heterogenisation of the the aminohydroxylation aminohydroxylation reaction reaction had had shown shown leaching leaching ranging ranging from from 20 20 to to 50% 50% [12]. [12]. heterogenisation of 2.4. Recycling Test [12]. heterogenisation of the aminohydroxylation reaction had shown leaching ranging from 20 to 50% [12]. [12]. 2.4. Test [12]. 2.4.Recycling Test One ofRecycling the main advantages of heterogeneous over homogeneous catalysis is that the catalyst can 2.4. Recycling Test 2.4. Recycling Test 2.4. Recycling Recycling Test be (easily) recycled a number of times [25].heterogeneous The results for over the recycling test showed thatisthe reaction One the 2.4. Test Oneof themain mainadvantages advantagesof heterogeneousover overhomogeneous homogeneouscatalysis catalysisisisthat thatthe thecatalyst catalyst 2.4. Recycling Test One ofofthe the main advantages ofofthe heterogeneous homogeneous catalysis that the catalyst time increases slightly with an increase in number of recycles (Table 3). One of main advantages of heterogeneous over homogeneous catalysis is that the catalyst can be (easily) recycled a anumber ofof times [25]. The results for the recycling test showed that the One of the main advantages of heterogeneous over homogeneous catalysis is that the catalyst can be (easily) recycled number times [25]. The results for the recycling test showed that the can be be (easily) recycled a number numberofof ofheterogeneous times [25]. [25]. The Theover results for the the recycling recycling test showed that the One of therecycled main advantages homogeneous catalysis is showed that the that catalyst can (easily) a times results for test the One of the main advantages of heterogeneous over homogeneous catalysis is that the catalyst reaction time increases slightly with an increase in the number of recycles (Table 3). can be (easily) (easily) recycledslightly number ofan times [25].inThe The results for the recycling test showed that that the the reaction timeincreases increases slightly with an increase inthe the number recycles (Table 3). reaction time with increase number ofofthe recycles (Table 3). can be recycled aa number of times [25]. results for recycling test showed reaction time increases with increase the number of recycles (Table 3). can be (easily) recycledslightly a number ofan times [25].in The results for the recycling test showed that the reaction time increases slightly with an increase in the number of recycles (Table 3). reaction time increases slightly with an increase in the number of recycles (Table 3). reaction time increases slightly with an increase in the number of recycles (Table 3).

Catalysts Catalysts2018, 2018,8,8,547 x FOR PEER REVIEW

77 of of 13 12

Table 3. Reaction time and percentage yield of the β-amino alcohols of cyclohexene with respect to Table 3. Reaction time and percentage yield of the β-amino alcohols of cyclohexene with respect to the the cycle number. cycle number.

Recycle Number Recycle Number 1 1 2 2 33

Reaction Time/h Conversion/Yield % * Reaction Time/h Conversion/Yield % * 7 >99 7 >99 7.5 >99 7.5 >99 8.5 >99 8.5 >99 ** determined byHPLC. HPLC. determined by

2.5. Spent Spent Catalyst Catalyst Characterisation Characterisation 2.5. The XRD XRD diffractogram diffractogram of of the The the spent spent catalyst catalyst (Figure (Figure 5) 5) shows showsthat thatthe thecatalyst catalystisisstill stillindeed indeeda hydrotalcite-like compound after the aminohydroxylation reaction. The low count in the XRD a hydrotalcite-like compound after the aminohydroxylation reaction. The low count in the XRD diffractogram could be the result of the spent catalyst losing some of its crystallinity during the course diffractogram could be the result of the spent catalyst losing some of its crystallinity during the course ofthe thereaction. reaction.ICP ICPanalysis analysis(Table (Table showed amount of osmium decreased of 4)4) showed thethe amount of osmium andand zinczinc hadhad decreased afterafter the the reaction, which indicates leaching from the catalyst structure to the reaction mixture. reaction, which indicates leaching from the catalyst structure to the reaction mixture.

◦ C and chloramine-T as Figure Figure 5.5. XRD XRD of of Os–HTlc Os–HTlc used used in in the the aminohydroxylation aminohydroxylation reaction reaction at at 25 25 °C and chloramine-T as nitrogen nitrogensource. source.

Table 4. The metal ratio in the catalyst before and after the reaction, as determined by ICP-OES analysis. Table 4. The metal ratio in the catalyst before and after the reaction, as determined by ICP-OES analysis. Catalyst Os Zn Al Before reaction Catalyst After reaction

3. Discussion

0.31Os 0.22

Before reaction After reaction

0.31 0.22

3.70 Zn 3.50

3.70 3.50

1.00 Al 1.00 1.00 1.00

3. Discussion The average distance between the cations (in the brucite-like structure) can be represented by the a parameter, calculated using the d-spacing of the in the HT and thecan HTlc’s doublet peak The average distance between the cations (infirst the peak brucite-like structure) be represented by (the 110 plane). For catalyst A, the a parameter (Table 1), which is similar to that presented in the the a parameter, calculated using the d-spacing of the first peak in the HT and the HTlc’s doublet literature found be 3.06A, Å.the Thea cparameter dimension(Table corresponds to thrice the 003 plane d-spacing peak (the[19], 110 was plane). For to catalyst 1), which is similar to that presented in (Table 1) and is also in agreement with that reported in literature [19]. the literature [19], was found to be 3.06 Å . The c dimension corresponds to thrice the 003 plane dThe (Table solvent demonstrated shorter reaction times with MeCN/H2 O and spacing 1) effect and is investigation also in agreement with that reported in literature [19]. t-BuOH/H v/v) than toluene. Taking into consideration the dielectric of the2β-amino 2 O (1:1 effect The solvent investigation demonstrated shorter reaction times constants with MeCN/H O and talcohol, usually around 25 [26], the aminohydroxylation reaction is best carried out in a system BuOH/H2O (1:1 v/v) than toluene. Taking into consideration the dielectric constants solvent of the β-amino that exhibits a high dielectric constant. MeCN/water (1:1 v/v) and t-BuOH/water (1:1 v/v) exhibit alcohol, usually around 25 [26], the aminohydroxylation reaction is best carried out in a solvent

system that exhibits a high dielectric constant. MeCN/water (1:1 v/v) and t-BuOH/water (1:1 v/v)


8 of 13

a higher dielectric constant than toluene (55.69 at 25 ◦ C and 47.94 at 60 ◦ C [26], and 36.59 at 20 ◦ C and 27.80 at 60 ◦ C [27], respectively, vs. 2.38 at 25 ◦ C for toluene [20]). It is thus evident why both MeCN/water (1:1 v/v) and t-BuOH/water (1:1 v/v) exhibit a shorter reaction time than toluene. Longer reaction times may also be observed for toluene as a solvent system, because water hydrolysis enhances the formation of the β-amino alcohol and, unlike MeCN and t-BuOH, water is immiscible with toluene. The versatility of the Os–HTlc catalyst in the dihydroxylation reaction with different olefins has been studied [19], showing 99.9% selectivity to diols, and it was found to be a truly heterogeneous system. The “ideal” composition of the Os–HTLc catalyst has also been investigated [20]. It was found that the Os/Zn/Al composition gives the most active catalyst in the dihydroxylation of cyclohexene, in terms of reaction time (for complete conversion of cyclohexene). The current results show that the Os–HTlc is active in the aminohydroxylation reaction with very high selectivity to the β-amino alcohols (~99.9%) of different olefin classes. The different classes of olefins only influence the reaction time (to complete olefin conversion), where functionalized olefins (Table 2, entry 4–7) required ~24 h reaction time. This could be the result of steric hindrance around the double bond limiting the interaction of the olefin and catalyst. This steric hindrance effect was also observed in the dihydroxylation reaction [19]. The results obtained showed slightly shorter reaction times for catalyst B, probably because of the much higher surface area (Table 1), which compensates for the lower Os content. However, catalyst B gave significantly lower selectivity to the β-amino alcohols, with a rise in diol formation. This could be attributed to the loss of the HTlc structure. The heterogeneity of the aminohydroxylation over Os–HTlc was investigated in term of Os leaching and catalyst recyclability under different reaction conditions. The solvent system influenced the Os leaching (in inactive form) level. This could be attributed to the chloramine-T solubility and its polarization effect, and/or the small amount of leached Os is the result of surface Os that was not fully incorporated in the HTlc structure. It is noteworthy that the leached Os is inactive and the aminohydroxlation reaction only occurs over the Os–HTlc catalyst. Reusability of the catalyst was a key aspect of this study in order to minimize the economic and environmental cost of Os on the process. Table 3 shows the reaction time (to complete olefin conversion) increased slightly from the first to third cycle, with no change to the β-amino alcohol yield (i.e., >99%) as a function of cycle number. This slight effect may be the result of the leaching of osmium or, more likely, a slow loss of HTlc structure. Some catalysts with a “better” HTlc structure are known to react faster than those with defects, as demonstrated in the dihydroxylation reaction [20]. To investigate the impact of the aminohdroxylation reaction on the Os–HTlc catalyst, the spent catalyst was characterized by XRD and BET surface area measurement, to study phase changes/crystallinity and surface area, respectively. The XRD results (Figure 5) show that the HTlc structure of the catalyst is maintained after the reaction, however, the crystallinity of the catalyst dropped. The BET surface area analyses of the spent catalyst showed an increase in the surface area of the catalyst to 107 m2 /g, from 87 m2 /g before the reaction (i.e., fresh catalyst). This change may be the result of grinding of the catalyst to a fine powder by the magnetic stirrer bar during the course of the reaction and/or redox processes on the catalyst’s surface causing a potential loss of HTlc structure. 4. Materials and Methods 4.1. Catalyst Preparation The preparation of the catalyst was carried out using the co-precipitation technique following a published method [12,13]. The salts OsCl3 ·nH2 O (0.53 g, 1.53 mmol), ZnCl2 (2.12 g, 15.3 mmol), and AlCl3 ·6H2 O (1.23 g, 5.10 mmol) were dissolved in 10 mL of de-ionized water. Sodium hydroxide solution 1M (46 mL) was used to dissolve sodium bicarbonate (1.41 g, 13.3 mmol). The two solutions (i.e., metals solution and base solution) were added drop-wise simultaneously into a 500 mL three necked round bottom flask, while the pH was maintained between 8 and 10. After complete addition


9 of 13

of the solutions, the resulting mixture was heated to 65 ± 5 ◦ C for 18 h. The solution was then allowed to cool to room temperature, after which the precipitate was filtered and washed to neutrality by large amounts of de-ionized water. Thereafter, the precipitate was dried in the oven for 12 h at 110 ◦ C (catalyst A). Catalyst B was prepared by heat treating a sample of catalyst A under a nitrogen atmosphere at 300 ◦ C for 2 h. 4.2. Catalyst Characterization The metal ratio was determined by inductively coupled plasma optical emission spectroscopy (ICP-OES) utilizing a Perkin Elmer Precisely Optical Emission Spectrometer, Optima 5300 DV (PerkinElmer, Shelton, CT, USA). Powder XRD was carried out using a Bruker D8 Advance (Karlsruhe, Germany) with diffracplus XRD commander software (version 2.5, Bruker AXS, Billerica, MA, USA), and a Bruker VANTEC detector (Bruker, Karlsruhe, Germany). The radiation source used was Cu Kα (wavelength of 0.1540 nm), operating on a long focus line with a voltage and amperage of 40 kV and 40 mA, respectively. The surface area of the different catalysts was obtained using a Micromeritics Gemini instrument (Micrometrics, Atlanta, GA, USA). Infrared analyses of the catalysts were carried out using a Perkin Elmer Precisely, Universal Attenuated Total Reflection (ATR) spectrum 100 series (PerkinElmer, Shelton, CT, USA). The geometry around aluminum was determined by 27 Al solid state nuclear magnetic resonance (SS-NMR), utilizing a Bucker 600 MHz Ultrashield instrument (Bruker, Karlsruhe, Germany), equipped with a 4 mm MAS BB/IH probe (Bruker, Karlsruhe, Germany) that was spinning at a speed of 12 kHz, with 1 s recycling speed. The dispersion of osmium in the catalyst was determined by scanning electron microscopy (SEM)–electron dispersion spectroscopy (EDS), utilizing a Jeol JSM-6100 scanning microscope (Joel, Peabody, MA, USA) with a Bruker signal processing unit detector (Bruker, Karlsruhe, Germany). 4.3. Catalytic Activity 4.3.1. Standard Catalytic Testing Method The aminohydroxylation reaction was carried out by adding 6 mL of solvent to a N2 filled Schlenk tube, followed by the addition of the olefin (0.478 mmol), nitrogen source (chloramine-T) (2 eq, 0.2173 g), and the catalyst (0.03 g), in this respective order. The reaction was monitored by a Perkin Elmer Autosystem XL gas chromatograph (PerkinElmer, Shelton, CT, USA) with a flame ionization detector (GC-FID), equipped with a CB Sil5 column, until complete depletion of the starting material (olefin). The β-amino alcohols have high boiling points; therefore, the products were analysed by a Shimadzu LC-8A high performance liquid chromatograph (Shimadzu, Kyoto, Japan) (HPLC) equipped with a C8 column and an ultra-violet detector. Each aminohydroxylation product was isolated and fully characterised (1 H and 13 C-NMR, IR, and mass spectrometry (MS)). Each reaction was repeated a minimum of two times. 4.3.2. Investigation into the Effect of the Solvent on the Reaction This study was carried out using three different solvents with different polarities. The solvents were toluene (6 mL toluene and 172 µL water), t-butanol/water (1:1 v/v), and acetonitrile/water (1:1 v/v). The olefin used in this study was cyclohexene and catalyst A was utilised as the respective catalyst. The catalytic testing was carried out as described in Section 4.3.1. 4.3.3. Investigation into the Effect of the Temperature on the Reaction This investigation was carried out as described in Section 4.3.1, employing three different temperatures (unless otherwise mentioned), namely, 25 ◦ C, 40 ◦ C, and 60 ◦ C. The olefin used in this part of the study was styrene and catalyst A was utilised as the respective catalyst.


10 of 13

4.3.4. Investigation into the Effect of the Catalyst Structure on the Reaction This investigation was carried out as described in Section 4.3.1., with catalysts A and B. The olefins used in this study were cyclohexene and t-butyl crotonate. The solvent utilized was acetonitrile/water (1:1 v/v), as described in Section 4.3.1. 4.3.5. β-Amino Alcohols Product Characterization N-(2-hydroxycyclohexyl)-4-methylbenzenesulfonamide (β-amino alcohol of cyclohexene, 23% and 25% isolated yield in t-BuOH/H2 O and MeCN/H2 O, respectively) 1H

NMR (400 MHz, CDCl3 ): δ 1.5–1.72 (m, 8H), 2.44 (s,3H), 3.21 (m, 1H), 3.78 (d, J = 2.48 Hz, 1H), 4.80 (d, J = 7.48 Hz, 1H), 7.32 (d, J = 8.04 Hz, 2H), 7.79 (d, J = 8.24 Hz,2H). 13 C NMR (100 MHz, CDCl3 ) δ 19.7 (s,1C), 21.5 (s,1C), 23.3 (s,1C), 27.9 (s,1C), 31.4 (s,1C), 55.1 (s,1C), 68.7 (s,1C), 126.9 (s,2C), 129.7 (s,2C), 137.9 (s,1C), 143.3 (s,1C). IR (cm−1 ) = 3414 (m), (OH), 3137 (m), (NH), 2938 (w), (CH2 ), 2849 (w), (CH3 ), 1598 (m), (Ar), 1029 (m), (S=O). Mass calculated = 269, MS = 291 m/z (M + Na). N-(hydroxyhexan-2-yl)-4-methylbenzenesulfonamide (β-amino alcohol of hexene, 8% and 8% isolated yield in t-BuOH/H2 O and MeCN/H2 O, respectively)

1H

NMR (400 MHz, CDCl3 ) δ 0.87 (t, J = 6.86 Hz, 3H), 1.2–1.40 (m, 6H), 2.42 (s, 3H), 2.7–2.80 (m, 1H), 3.0–3.01 (m, 1H), 3.6–3.70 (m, 1H), 7.31 (d, J = 8.04 Hz, 2H), 7.74 (d, J = 8.20 Hz, 2H). 13 C NMR (100 MHz, CDCl3 ) δ 13.9 (s,1C), 21.5 (s,1C), 22.5 (s,1C), 27.4 (s,1C), 34.3 (s,1C), 48.6 (s,1C), 70.4 (s,1C), 127.1 (s,2C), 129.7 (s,2C), 136.7 (s,1C), 143.5 (s,1C). Mass calculated = 271.1, MS = 293.1 m/z (M + Na). N-(-2hydroxyl-1-phenylethyl)-4-methylbenzenesulfonamide (β-amino alcohol of Styrene, 13% and 14% isolated yield in t-BuOH/H2 O and MeCN/H2 O, respectively) 1H

NMR (400 MHz, CDCl3 ) δ 2.42 (s, 3H), 3.0–3.03 (m, 1H), 3.2–3.29 (m, 1H), 4.79 (d, J = 4.40 Hz, 1H), 4.89 (d, J = 4.40 Hz, 1H), 7.2–7.34 (m, 7H), 7.72 (d, J = 8.28 Hz, 2H). 13 C NMR (100 MHz, CDCl3 ) δ 21.5 (s,1C), 30.9 (s,1C),50.1 (s,1C), 125.8 (s,2C), 127.1 (s,2C), 128.3 (s,1C), 128.7 (s,2C), 129.8 (s,2C), 136.7 (s,1C), 140.7 (s,1C), 143.6 (s,1C). IR (cm−1 ) =3399 (m), (OH), 3149 (m), (NH), 2926 (w), (CH2 ), 2862 (w), (CH3 ), 1599 (w), (Ar), 1086 (m), (S=O). Mass calculated = 291, MS = 313 m/z (M + Na). N-(2-hydroxyl-1,2-diphenylethyl)-4-methylbenzenesulfonamide (β-amino alcohol of cis-stilbene, 30% and 30% isolated yield in t-BuOH/H2 O and MeCN/H2 O, respectively) 1H

NMR (400 MHz, DMSO) δ 2.21 (s, 3H), 4.28 (dd, J = 6.80, 9.20 Hz, 1H), 4.62 (dd, J = 6.80, 9.20 Hz, 1H), 5.37 (d, J = 4.80 Hz, 1H), 7.0–7.29 (m, 14H), 8.08 (d, J = 4.80 Hz, 1H). 13 C NMR (100 MHz, DMSO) δ 20.8 (s,1C), 63.3 (s,1C), 75.3 (s,1C), 126.1(s,2C), 126.3 (s,1C), 126.7 (s,2C), 126.9 (s,1C), 127.0 (s,2C), 127.5 (s,2C), 128.2 (s,2C), 128.8 (s,2C), 138.5 (s,1C), 138.8 (s,1C), 141.6 (s,1C), 142.6 (s,1C). IR (cm−1 ) = 3461 (m), (OH), 3323(m), (NH), 3031 (w), 1598 (w), (Ar), 1055 (m), (S=O). Mass calculated = 367.1, MS = 389.12 m/z (M + Na). Methyl-2-hydroxy-3-(4-methylbenzenesulfonamide)-3-phenylpropanoate (β-amino alcohol of methylcinnamate, 35% and 34% isolated yield in t-BuOH/H2 O and MeCN/H2 O, respectively) 1H

NMR (400 MHz, CDCl3 ) δ 2.32 (s, 3H), 3.18 (d, J= 4.16 Hz,1H), 3.75 (s, 3H), 4.34 (dd, J = 2.36, 4.04 Hz, 1H), 4.85 (dd, J = 2.22, 9.70 Hz, 1H), 5.52 (d, J = 9.72 Hz, 1H), 7.0–7.18 (m, 7H), 7.53 (d, J = 8.20 Hz, 2H). 13 C NMR (100 MHz, CDCl3 ) δ 21.4 (s,1C), 53.2 (s,1C), 58.9 (s,1C), 74.2 (s,1C), 126.8 (s,2C), 126.9 (s,2C), 127.8 (s,2C), 128.4 (s,2C), 129.2 (s,1C), 137.4 (s,1C), 137.5 (s,1C), 143.1 (s,1C), 172.4 (s,1C). IR (cm−1 ) = 3477 (m), (OH), 3139 (m), (NH), 2967 (w), (CH3 ), 2882 (w), (CH3 ), 1598(w), (Ar), 1738 (m), (C=O), 1056 (m), (S=O). Mass calculated = 349.3, MS = 372 m/z (M + Na). Dimethyl-2-hydroxy-3-(4-methylbenzenesulfonamide)succinate (β-amino alcohol of dimethylfumarate, 30% and 31% isolated yield in t-BuOH/H2 O and MeCN/H2 O, respectively)


11 of 13

1H

NMR (400 MHz, CDCl3 ) δ 2.41 (s, 3H), 3.23 (d, J = 5.20 Hz, 1H), 3.60 (s, 3H), 3.76 (s, 3H), 4.39 (dd, J = 1.92, 9.80 Hz, 1H), 4.59 (dd, J = 1.78, 5.02 Hz, 1H), 5.37 (d, J = 9.88 Hz, 1H), 7.29 (d, J = 8.20 Hz, 2H), 7.70 (d, J = 8.28 Hz, 2H). 13 C NMR (100 MHz, CDCl3 )_21.5 (s,1C), 53.1 (s,1C), 53.4 (s,1C), 57.9 (s,1C), 71.5 (s,1C), 127.2 (s,2C), 129.6 (s,2C), 136.5 (s,1C), 143.8 (s,1C), 168.7 (s,1C), 171.3 (s,1C). IR (cm−1 ) = 3503 (m), (OH), 3260 (m), (NH), 2923 (w), 1598 (w), (Ar), 1738 (m), 1730 (m), (C=O), 1056 (m), (S=O). Mass calculated = 331, MS = 353 m/z (M + Na). tert-Butyl-2-hydroxy-3-(4-methylbenzenesulfonamide)butanoate (β-amino alcohol of tert-butylcrotonate, 32% and 32% isolated yield in t-BuOH/H2 O and MeCN/H2 O, respectively)

1H

NMR (400 MHz, CDCl3 ) δ 0.98 (d, J = 6.72 Hz, 3H), 1.51 (s, 9H), 2.41 (s, 3H), 3.22 (d, J = 3.64 Hz, 1H), 3.8–3.86 (m, 2H), 4.75 (d, J = 10.08 Hz, 1H),7.29 (d, J = 7.96 Hz, 2H), 7.75 (d, J = 8.32 Hz, 2H). 13 C NMR (100 MHz, CDCl3 ) δ 17.9 (s,1C), 21.5 (s,1C), 27.9 (s,3C), 51.5 (s,1C), 73.6 (s,1C), 84.1 (s,1C), 126.9 (s,2C), 129.7 (s,2C), 138.6 (s,1C),143.3 (s,1C), 171.6 (s,1C). IR (cm−1 ) = 3446 (m), (OH), 3260 (m), (NH), 2985 (w), (Ar), 2919 (w), (CH3 ), 1598 (w), (ar), 1716 (m), (C=O), 1048 (m), (S=O). Mass calculated = 329, MS = 351 m/z (M + Na). 4.4. Leaching Test This study was carried out in two solvent systems, MeCN/water (1:1 v/v) and t-BuOH/water (1:1 v/v) (Section 4.3.1). The catalyst used was catalyst A and cyclohexene was the olefin. However, the reaction was stopped after approximately 20% conversion of cyclohexene (3 h). Thereafter, the catalyst was removed by gravity filtration and the reaction mixture was centrifuged at 3000 rpm for 15 min, for the removal of any suspended fine catalyst particles in the reaction mixture. The reaction solution was then stirred until the end of the normal duration that the reaction takes in the presence of the catalyst. At the end of this period, GC analysis was undertaken to investigate if any further conversion of the starting material had occurred in the absence of the catalyst, as well as to determine if any osmium that may have leached was catalytically active or not. Furthermore, to determine if any osmium leached from the catalyst, ICP-OES analysis of the solution was undertaken. 4.5. Recycling Test For the determination of the reusability of the catalyst, it was recycled three times. The reaction was carried out as described in Section 2.3. At the end of the reaction, the catalyst was filtered by vacuum filtration and reused with fresh starting material, that is, olefin (cyclohexene), nitrogen source (chloramine-T) (2 eq), and solvent system (i.e., MeCN/water (1:1 v/v). 5. Conclusions The aminohydroxylation reaction can be carried out heterogeneously using hydrotalcite-like catalysts containing osmium. Importantly, this catalyst also gave a very high selectivity to the aminohydroxylation products and, consequently, a very low selectivity to the diols, unlike other catalysts in literature. Indeed, only trace quantities of diol were observed. Functionalised and electron deficient olefins showed the longest reaction times. Characterisation of the spent catalyst showed that the catalyst maintained the hydrotalcite-like structure after the reaction. This study has shown the lowest amount of osmium leach reported yet. Furthermore, it is possible that the use of a different nitrogen source may eliminate leaching altogether. We further suggest that using a chiral nitrogen source or ligand may allow for an asymmetric reaction. Author Contributions: M.I.F. carried out all the experimental work and original manuscript preparation. G.E.M.M. and H.B.F. acted in supervision roles, validating and reviewing the manuscript. Funding: This research was funded by Mintek (AMI program) and the South African Department of Science and Technology (DST). Acknowledgments: The Electron Microscopy Unit at UKZN, Westville is thanked for the SEM and SEM–EDS analyses.


12 of 13

Conflicts of Interest: The authors report no conflict of interest.

References 1. 2. 3. 4. 5. 6. 7. 8. 9.

10. 11. 12. 13.

14. 15. 16. 17. 18. 19. 20. 21. 22. 23. 24.

25.

Kolb, H.C.; Sharpless, K.B. Asymmetric aminohydroxylation. In Transition Metals for Organic Synthesis; Matthias Beller, C.B., Ed.; Wiley-VCH: Weinheim, Germany, 1998. Kolb, H.C.; VanNieuwenhze, M.S.; Sharpless, K.B. Catalytic asymmetric dihydroxylation. Chem. Rev. 1994, 94, 2483–2547. [CrossRef] Dupau, P.; Epple, R.; Thomas, A.A.; Fokin, V.V.; Sharpless, K.B. Osmium-catalyzed dihydroxylation of olefins in acidic media: Old process, new tricks. Adv. Synth. Catal. 2002, 344, 421–433. [CrossRef] Li, G.; Chang, H.-T.; Sharpless, K.B. Catalytic asymmetric aminohydroxylation (aa) of olefins. Angew. Chem. Int. Ed. Engl. 1996, 35, 451–454. [CrossRef] Li, G.; Angert, H.H.; Sharpless, K.B. N-halocarbamate salts lead to more efficient catalytic asymmetric aminohydroxylation. Angew. Chem. Int. Ed. Engl. 1996, 35, 2813–2817. [CrossRef] Nilov, D.; Reiser, O. The sharpless asymmetric aminohydroxylation—Scope and limitation. Adv. Synth. Catal. 2002, 344, 1169–1173. [CrossRef] Bergmeier, S.C. The synthesis of vicinal amino alcohols. Tetrahedron 2000, 56, 2561–2576. [CrossRef] Lohray, B.B.; Thombare, P.S.; Bhushan, V. Advances in asymmetric aminhydroxylation. PINSA 2002, 68A, 391. Kattamuri, P.V.; Salmonsen, R.; McQuain, C.; Burstein, S.; Sun, H.; Li, G. Asymmetric synthesis of novel n-(1-phenyl-2,3-dihydroxypropyl)arachidonylamides and evaluation of their anti-inflammatory activity. Life Sci. 2013, 92, 506–511. [CrossRef] [PubMed] Williamson, K.S.; Yoon, T.P. Iron-catalyzed aminohydroxylation of olefins. J. Am. Chem. Soc. 2010, 132, 4570–4571. [CrossRef] [PubMed] Michaelis, D.J.; Shaffer, C.J.; Yoon, T.P. Copper(ii)-catalyzed aminohydroxylation of olefins. J. Am. Chem. Soc. 2007, 129, 1866–1867. [CrossRef] [PubMed] Song, C.; Oh, C.; Lee, S.; Sherrington, D. Heterogeneous asymmetric aminohydroxylation of alkenes using a silica gel-supported bis-cinchona alkaloid. Chem. Commun. 1998, 22, 2435–2436. [CrossRef] Mandoli, A.; Pini, D.; Agostini, A.; Salvadori, P. Heterogeneous asymmetric catalytic aminohydroxylation promoted by a bis-cinchona alkaloid derivative supported onto an insoluble organic polymeric matrix. Tetrahedron Asymmetry 2000, 11, 4039–4042. [CrossRef] Choudary, B.M.; Chowdari, N.S.; Jyothi, K.; Kantam, M.L. Heterogeneous catalytic asymmetric aminohydroxylation of olefins using ldh-supported oso4. J. Mol. Catal. A Chem. 2003, 196, 151–156. [CrossRef] Vaccari, A. Preparation and catalytic properties of cationic and anionic clays. Catal. Today 1998, 41, 53–71. [CrossRef] Cavani, F.; Trifirò, F.; Vaccari, A. Hydrotalcite-type anionic clays: Preparation, properties and applications. Catal. Today 1991, 11, 173–301. [CrossRef] Othman, M.R.; Helwani, Z.; Fernando, W.J.N. Synthetic hydrotalcites from different routes and their application as catalysts and gas adsorbents: A review. Appl. Organomet. Chem. 2009, 23, 335–346. [CrossRef] Oswald, H.R.; Asper, R. Physics and Chemistry of Materials with Layered Structures; Reidel: Dordrecht, The Netherlands, 1977; Volume 73. Friedrich, H.B.; Govender, M.; Makhoba, X.; Ngcobo, T.D.; Onani, M.O. The os/cu-al-hydrotalcite catalysed hydroxylation of alkenes. Chem. Commun. 2003, 23, 2922–2923. [CrossRef] Naicker, T.; Datye, A.K.; Friedrich, H.B. A comparative study of os-hydrotalcites for the cis-dihydroxylation of cyclohexene. Appl. Catal. A Gen. 2008, 350, 96–102. [CrossRef] Miyata, S. Hydrotalcites in relation to composition. Clays Clay Miner. 1980, 28, 50–56. [CrossRef] Kosower, E.M. The effect of solvent on spectra. I. A new empirical measure of solvent polarity: Z-values. J. Am. Chem. Soc. 1958, 80, 3253–3260. [CrossRef] Sheldon, R.A.; Wallau, M.; Arends, I.W.C.E.; Schuchardt, U. Heterogeneous catalysts for liquid-phase oxidations: Philosophers’ stones or trojan horses? Acc. Chem. Res. 1998, 31, 485–493. [CrossRef] Yang, X.-W.; Liu, H.-Q.; Xu, M.-H.; Lin, G.-Q. A highly efficient and practical new peg-bound bi-cinchona alkaloid ligand for the catalytic asymmetric aminohydroxylation of alkenes. Tetrahedron Asymmetry 2004, 15, 1915–1918. [CrossRef] Hagen, J. Industrial Catalysis: A Practical Approach; Wiley-VCH: Weinheim-Hohensachsen, Germany, 2006.


26.

27.

13 of 13

Gagliardi, L.G.; Castells, C.B.; Ràfols, C.; Rosés, M.; Bosch, E. Static dielectric constants of acetonitrile/water mixtures at different temperatures and debye−hückel a and a0b parameters for activity coefficients. J. Chem. Eng. Data 2007, 52, 1103–1107. [CrossRef] Akerlof, G. Dielectric constants of some organic solvent-water mixtures at various temperatures. J. Am. Chem. Soc. 1932, 54, 4125–4139. [CrossRef] © 2018 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/).