Anuj K.
Rathi
a,
Manoj B.
Gawande
*a,
Vaclav
Ranc
a,
Jiri
Pechousek
a,
Martin
Petr
a,
Klara
Cepe
a,
Rajender S.
Varma
b and
Radek
Zboril
*a
aRegional Centre of Advanced Technologies and Materials, Department of Physical Chemistry, Faculty of Science, Palacky University, Šlechtitelů 11, Olomouc, 783 71, Czech Republic. E-mail: manoj.gawande@upol.cz; radek.zboril@upol.cz
bSustainable Technology Division, National Risk Management Research Laboratory, US Environmental Protection Agency, 26 West Martin Luther King Drive, MS 443, Cincinnati, Ohio 45268, USA
First published on 7th September 2015
Maghemite-supported ultra-fine Pd (1–3 nm) nanoparticles, prepared by a simple co-precipitation method, find application in the catalytic continuous flow hydrogenation of nitroarenes, azides, and alkenes wherein they play an important role in the reduction of various functional groups on the surface of maghemite with catalyst loading (~6 wt% Pd). The salient features of the protocol include expeditious formation of reduced products in high yields under near ambient conditions with recycling of the catalyst (up to 12 cycles) without any decrease in selectivity and yield.
Continuous flow reactions are receiving tremendous interest for applications in the development of selective processes because of their intrinsic advantages, when compared to conventional batch reactions; the ease of isolation or purification, reduction in waste emissions, safety, automation, and space-time-yield efficiency being prominent.5
The chemoselective and partial catalytic hydrogenation of functionalized hydrocarbons with multiple CC and CC bonds is a very desirable process and prerequisite in the pharmaceutical and petrochemical industries.6 In general, the catalytic hydrogenation reaction (reduction of nitro compounds, CC and CC bonds) often deploys various types of heterogeneous and noble metal-supported catalysts under continuous flow conditions.7 However, in most cases, catalysts cannot be reused and recycled or plagued by leaching of metals due to high pressure and related reaction parameters in the flow reactor.8 Consequently, it is prudent to design a stable, cost-effective, and reusable catalytic system for hydrogenation reactions.
In recent years, there has been a remarkable rise in the exploitation of magnetic nanocomposites because of their preparation from inexpensive precursors, inert and stable nature and most importantly, their reusable and recyclable feature for several runs without impairment of catalytic activity and selectivity.9 It is not surprising, therefore, that there is an enormous demand for magnetically recyclable nanocatalysts in continuous flow processes.5c Herein, we describe hydrogenation reactions for the reduction of nitro, azide, and alkene functionalities under a continuous flow on recyclable maghemite decorated with Pd nanoparticles (Fig. 1).
The XRD patterns of the maghemite and maghemite–Pd nanocomposite are shown in Fig. 2. In both profiles, all the peaks correspond to the maghemite's (γ-Fe2O3) structure; the presence of reflections at (110), (210), and (211) supports the occurrence of maghemite with partially ordered vacancies (PDF card 01-089-5892). Furthermore, Rietveld analysis was performed on both XRD patterns to calculate the lattice parameter of maghemite (a = 8.355 Å) and the average crystalline domain size is calculated using the broadening of the most intense peaks, implying the crystallite sizes of the samples in nanometer dimensions (~15 nm). Palladium-related diffraction peaks are barely extractable from the XRD pattern of maghemite–Pd nanocomposite which can be attributed either to its low percentage or the much reduced size of the Pd particles (<3 nm).10
Fig. 3 shows the TEM image of the maghemite–Pd catalyst with a nearly spherical geometry of the maghemite support having a size in the range of 10–30 nm, thus corroborating the average size determination of maghemite from XRD. The maghemite nanoparticles are homogeneously covered by Pd nanoparticles – the fact explaining the absence of Pd peaks in the XRD pattern of maghemite–Pd nanocomposite. The measurement of the size of Pd NPs by HRTEM clearly indicates the presence of ultra-small Pd nanoparticles (1–3 nm) (Fig. 3c and d). Furthermore, the morphology of maghemite–Pd after 12 consecutive cycles was verified by TEM; no major visible changes in the morphology were observed (Fig. S1b, ESI†).
To explore the chemical nature of palladium in the maghemite–Pd sample, XPS analysis was performed for the freshly prepared maghemite–Pd and the reused catalyst after 12 reactions cycles (Fig. 4); the XPS spectra of both samples exhibit dominantly Pd(0) and PdO species without any significant change after recycling (Fig. 4). The most intense peaks of doublet in the maghemite–Pd sample at 335.50 eV and 340.76 eV are assignable to metallic Pd(0) and the second weaker set of peaks at 336.81 eV and 342.07 eV could be assigned to Pd(II) in an oxidized form such as PdO.11 Thus, the TEM data clearly manifest that the maghemite–Pd nanocatalyst is composed of globular maghemite (10–30 nm) nanoparticles. The palladium particles are predominantly in the form of metallic Pd with a minor presence of PdO.
Fig. 4 XPS spectrum of (a) maghemite–Pd nanocatalyst and (b) recycled maghemite–Pd after 12 reactions; positions of the metallic Pd(0) and PdO are denoted in blue and red, respectively. |
To distinguish chemical components of the nanocomposite and the chemical nature of the magnetic support (maghemite vs. magnetite), HAADF-STEM and Mössbauer spectroscopy were conducted. High-angle annular dark-field scanning transmission electron microscopy (Fig. 5) confirmed that Pd nanoparticles (green) functionalize the iron oxide surface very homogeneously.
Fig. 5 High-angle annular dark-field scanning transmission electron microscopy (HAADF-STEM) images showing elemental mapping of Pd, Fe, and O atoms. |
57Fe Mössbauer spectroscopy on maghemite–Pd was carried out to clarify the oxidation state of Fe ions in the synthesized and recycled samples as the XRD data cannot distinguish between maghemite and magnetite. The recorded room temperature/zero-field (300 K/0 T) and low temperature in-field Mössbauer spectra (5 K/5 T) are shown in Fig. 6 and the values of the Mössbauer hyperfine parameters, derived from the spectra fitting, are listed in Table S1, ESI.†
The room temperature Mössbauer spectra of the synthesized maghemite–Pd and the recovered maghemite–Pd samples show one asymmetrical sextet component (Fig. 6a and c). The profiles of the spectra are typical of the γ-Fe2O3 phase with iron cations solely in the Fe3+ oxidation state.12 The Mössbauer spectra were recorded at a low temperature (5 K) to facilitate the resolution of the Mössbauer resonant lines via application of an external magnetic field of 5 T. The Mössbauer spectra of both samples are well fitted with two sextets (Fig. 6b and d). The first sextet with a lower isomer shift (δ) and higher hyperfine magnetic field (Beff) corresponds to the Fe3+ ions in the tetrahedral sites (T-sites) of the spinel γ-Fe2O3 crystal structure while the second sextet with a higher δ and lower Beff is ascribable to the Fe3+ ions occupying the octahedral sites (O-sites) in the γ-Fe2O3 crystal structure.
For the synthesized sample, the spectral ratio between the T-sextet to O-sextet is very close to 0.6 (Fig. 6b) indicating a stoichiometric nature of maghemite with vacancies present only in the O-sites. However, after the reaction, for the recovered sample (Fig. 6d), the spectral ratio of T:O is below 0.5, which reflects a partially non-stoichiometric character of the maghemite particles, probably due to the occurrence of vacancies at the T sites. In summary, the in-field Mössbauer data confirm well the stoichiometric maghemite structure with no indications of Fe(II) ions. Similarly, after the sample recycling, there is no remarkable change in the structure/valence state of the magnetic support. This fact reflects, together with the previously discussed XPS data, the stability of maghemite–Pd nanocatalyst keeping its chemical nature/crystal structure after recycling. As per analysis, we observed that the hyperfine parameters slightly changed, which might be due to the small changes in the iron surroundings or palladium in the sample after the reaction.
The applicability of the maghemite–Pd nanocatalyst was evaluated in the reduction of nitroarenes using 4-methoxy nitrobenzene as the model substrate in ethanol at 30 °C (Table 1). First, the control reaction was conducted with only maghemite; in the absence of palladium, no product formation was observed (Table 1, entry 1).
Entry | Catalyst | Temp. | Conversionb (%) | Isolated yield (%) |
---|---|---|---|---|
a Reaction conditions: 4-methoxy nitrobenzene (1 mmol), maghemite–Pd (wt%), EtOH (5 mL), flow rate (500 uL min−1), H2 (gas flow rate 60 mL min−1). b Conversion calculated on the basis of GC analysis. c H2 (30 mL per minute). tr = 0.75 min. | ||||
1 | Maghemite | 70°C | — | — |
2 | Maghemite–Pd (9%) | 70°C | >99 | 94 |
3 | Maghemite–Pd (7%) | 70°C | >99 | 95 |
4 | Maghemite–Pd (7%) | 30°C | >99 | 93 |
5 | Maghemite–Pd (6.2%) | 30°C | >99 | 94 |
6 | Maghemite–Pd (6.2%) | 30°C | >81 | 72c |
7 | Maghemite–Pd (3.9%) | 30°C | >65 | 53 |
The optimized reaction conditions were identified by conducting a series of experiments with different Pd loadings and variation of temperature. The results revealed that full conversion could be achieved with ~6.2 wt% of Pd loadings using hydrogen in a full mode (full H2 mode which corresponds to a gas flow rate of 60 mL min−1) at a 500 uL min−1 flow rate (Table 1, entries 2–5), while 30 mL min−1 hydrogen pressure showed 81% conversion (Table 1, entry 6). It was noticed that the low loading of catalyst (3.92%) showed only 65% conversion (Table 1, entry 7). These optimized reaction conditions were then applied to the reduction of a variety of substrates bearing additional and reducible functional groups (–CONH2, –SMe, –CN, –OH, –SO2NH2, –COOH etc.) attached to the aromatic ring (Table 2).
Entry | Substrate | Product | Yieldb (%) |
---|---|---|---|
a Reaction conditions: nitro compound (1 mmol), maghemite–Pd (6.2 wt%), EtOH:EtOAc (5 mL, 1:1), temp 30 °C, flow rate (500 uL min−1), H2 (gas flow rate 60 mL min−1), tr = 0.75 min. b Isolated yield. c Isolated by crystallization. d Isolated as hydrochloride salt. e N-Benzyl-4-nitroaniline (0.5 mmol), maghemite–Pd (6.2 wt%), EtOAc:EtOH (12 mL, 5:1), temp 70 °C, 300 uL min−1, tr = 1.25 min. | |||
1 | 94 | ||
2 | 94 | ||
3 | 94 | ||
4 | 92 | ||
5 | 98c | ||
6 | 93d | ||
7 | 95c | ||
8 | 95 | ||
9 | 87 | ||
10 | 88c | ||
11 | 93 | ||
12 | 95 | ||
13 | 86e |
In all cases, excellent yields of amine (86–98%) were obtained by just passing the solution of nitroarenes in ethanol:ethyl acetate (1:1) through the catalyst cartridge (Table 2, entries 1–10 and 13), the only exceptions being 4-nitrobenzonitrile and 4-nitrobenzoic acid which afforded 87% and 88% isolated yields, respectively (Table 2, entries 9, 10). The use of ethyl acetate was found beneficial as it helps improve the solubility of nitroarenes. Among the acyclic compounds, nitrocyclopentane gave a high yield (93%) of cyclopentyl amine, with no detectable side products, although mechanistically, the formation of amine proceeds via the intermediacy of nitroso derivatives and hydroxylamine species;13 the formation of hydroxylamines during the catalytic hydrogenation is often implicated in the ensuing explosions.13 The azido functionality is often deployed in organic synthesis and usually serves as a latent amino group which can be converted into an amine via reduction. The optimized conditions were utilized for the reduction of 1-azido-4-methoxy benzene and 1-azido-4-methyl benzene, which afforded excellent yields of products of 93% and 95%, respectively (Table 2, entries 11 and 12). Similarly, the reduction of N-benzyl-4-nitroaniline gave debenzylated benzene-1-4-diamine in 86% yield (Table 2, entry 13).
The viability of the catalytic system was further explored for the reduction of several of substituted alkenes using a H2 gas flow rate of 60 mL min−1 (full mode) at 50 °C (Table 3). In this case, ethyl benzene, 3-amino ethyl benzene, 4-methoxy ethyl benzene and 1,2-diphenyl ethane were obtained in a nearly quantitative conversion, 4-fluro ethyl benzene and cyclooctene were obtained in >98% and >91% conversion, respectively. It was observed that when the flow rate is increased from 300 to 500 uL min−1, the reaction was incomplete and some unreacted starting material remained. Additionally, methyl cinnamate and methyl 3-(4-methoxyphenyl)acrylate underwent efficient hydrogenation with a flow rate of 300 uL min−1 at 70 °C (Table 3, entries 7 and 8).
Entry | Substrate | Product | Conversionb (%) | Yieldc (%) |
---|---|---|---|---|
a Reaction conditions: substrate (1 mmol), maghemite–Pd (6.2 wt%), EtOH (5 mL), temp. 50 °C, flow rate 300 uL min−1, H2 (gas flow rate 60 mL min−1), tr = 0.75 min. b Conversions calculated on the basis of GC analysis. c Isolated yield. d Yields are determined by GC against internal integration. e Temp. 70 °C, isolated yields. | ||||
1 | >99 | 95d | ||
2 | >98 | 93d | ||
3 | >99 | 96d | ||
4 | >91 | 86d | ||
5 | >99 | 94d | ||
6 | >99 | 93d | ||
7 | >99 | 94e | ||
8 | >99 | 95e |
The salient beneficial feature of the reduction under flow conditions has been the recyclability of the catalyst. It is important to note that this magnetic support (maghemite) plays a crucial role in the successful reusability of the catalysts. Due to the presence of a hydroxyl group on the surface of the maghemite support, there is a sturdy interaction between maghemite and Pd NPs, which certainly results in the minimal leaching of Pd. The developed catalyst was reused and recycled for 12 times successfully with excellent conversion and yield without any reactivation, which are important performance metrics for cost-effective industrial processes (see Table S2, ESI†). The ICP-MS analysis indicates 0.000126, 0.000121, and 0.000113 Pd% leaching after the first, third and seventh reaction cycles.
The hydrogenation reactions, catalyzed by maghemite–Pd nanocomposites, in a continuous flow mode have several advantages relative to batch processes, where unique solid–gas–liquid triphasic reaction conditions prevail during the hydrogenation reactions.14 Most of the reported traditional catalysts are deployed under homogeneous conditions and are not reusable, making the protocol expensive.15 Also, the present catalytic protocol is comparable with reported methods used in continuous flow methods (Tables S3 and S4, ESI†). In contrast, maghemite–Pd nanocomposites could be easily retrieved and reused for several reaction cycles with barely observable leaching of Pd metal, which renders this catalytic protocol truly sustainable. Additionally, no hydrogen balloon/gas cylinder is used for the reactions, as in situ generated hydrogen gas from water certainly increases the safety aspects.16 We believe that several other hydrogenation reactions could be performed using this catalyst.
Footnote |
† Electronic supplementary information (ESI) available: Materials and reagents used, description of characterization techniques employed, TEM images of fresh catalyst and reused catalyst, values of the Mössbauer hyperfine parameters derived from the spectra fitting, and conversion values after reusability. See DOI: 10.1039/c5cy00956a |
This journal is © The Royal Society of Chemistry 2016 |