1. Hydrogenation of Alkenes
Four Pd-doped (0.25–0.45 wt.%), sulfonated, synthetic IERs were tested in the liquid-phase hydrogenation of cyclohexene to cyclohexane and cyclohexen-1-one to cyclohexanone by Zecca and coworkers
[1]. All resins were exchanged with palladium acetate, and Pd
II was reduced to Pd
0 with NaBH
4 in ethanol. Depending on their composition, the solvent affinities of the resins spanned from moderately hydrophobic to clearly hydrophilic—the lower the ST amount in the resin, the higher its hydrophilicity. No effect of the hydrophobic/hydrophilic character of the resin was observed on the reported catalytic activity. Interestingly, resins with more nitrogen atoms in their structures (i.e., those based on
N,
N′-methylene(bis)acrylamide and
N,
N-dimethylacrylamide) were found to be the most active ones, with rates increasing linearly with the nitrogen/Pd molar ratio, and no metal leaching was observed in any case.
The role of the gel-type IERs properties in the hydrogenation of the C═C bonds of maleic and fumaric acids, viz. the
cis and
trans isomers of butenedioic acid, was investigated in Drelinkiewicz et al. (2008)
[2]. They synthesized Pd-doped (0.25–2.0 wt.%, 2–3 nm) acrylic resins with crosslinking degrees of 3 and 10%. The observed swelling of the polymer mass during the hydrogenation was linked to the catalytic activity: the same hydrogenation rates were observed for both maleic and fumaric acids in the absence of swelling, while the hydrogenation rate of maleic acid was much higher than that of fumaric acid in the swollen state. Interestingly, a “bell-like” dependence between initial rates and metal content was found, with maximum activity at 0.5–1.0 wt.% Pd content.
Among the reactions tested in Moreno Marrodan et al. (2012), hydrogenation of the C═C bond in methyl 2-acetamidoacrylate and in
trans-4-phenyl-3-buten-2-one was studied over Pd-doped, gel-type, cationic and anionic resins
[3]. Reported conversions were 91.7% and 98.2%, respectively, with selectivities of 100% and 83.7%, over lithiated Pd/Dowex
TM 50W × 2 resin without any separated pre-reduction step of the ionic Pd
II to Pd
0. Remarkably, neither activity loss upon recycling of the catalyst nor Pd leaching was observed and selectivity was unaffected after 6 cycles using methanol as the reaction solvent.
Madureira and associates hydrogenated fatty acid methyl esters (i.e., methyl undecenoate, methyl oleate, methyl ricinoleate, and methyl linoleate) and vegetable oil-based compounds (i.e., glycerol trioleate, castor oil, jojoba oil, olive oil, and very high-oleic sunflower oil) over Ru/Dowex
TM 50W × 2 under mild conditions (30 °C and 10 bar H
2)
[4]. Regarding the metal impregnation procedure (consisting of ion exchange with an aqueous metal salt solution followed by NaBH
4 reduction), the authors indicate that the resin must be in the Na
+ form to improve metalation yield and to avoid Ru leaching due to the presence of acidic species. According to their results, the catalytic activity of the catalysts was higher for shorter hydrocarbon chain length: hydrogenation of methyl undecenoate (C
12) was faster than methyl linoleate (C
19), showing 100% conversion in 60 min in front of about 50% conversion in 210 min, respectively. Moreover, the double bond in the larger unsaturated esters was located in an internal position, which hindered the reaction rate
[4].
In a different approach, a lithiated Dowex
TM 50W × 2 resin was used to immobilize phosphine–rhodium complexes to catalyze the asymmetric hydrogenation of prochiral olefins, such as methyl 2-acetamidoacrylate, under mild conditions (room temperature, under 1–5 bar H
2)
[5][6][7]. Reported yields mount up to 99.9%, even in the second recycle, with a detected Rh leaching of 2.0%
[5].
Hydrogenation of cyclohexene in supercritical CO
2 over metal-doped Amberlyst
TM 15 was studied by Seki and associates
[8]. The metals employed (Pd, Pt, Ru, and Rh) were incorporated into Amberlyst
TM 15 in different amounts from 0.5 to 2 wt.% to obtain monometallic catalysts. A bimetallic Pd-Pt catalyst was also tested. The highest conversions (i.e., up to 99%) and selectivity towards cyclohexane (i.e., >99%) were reported for Pd-based catalysts. Results obtained with Pt, Ru, and Rh catalysts were clearly less favorable with, conversion values in the range of 1–4%.
Some relevant hydrogenation reactions of alkenes found in the covered literature are shown in Scheme 1. Table 1 lists resin types, tested reactions, and obtained results of some of the reviewed works.
Scheme 1. Examples of hydrogenation reactions of alkenes over metal-doped IERs in the revised literature.
Table 1. Hydrogenation reactions of alkenes over metal-doped IERs in the literature.
As
Scheme 1 and
Table 1 reveal, Pd is the preferred metal to hydrogenate C═C bonds, as described in previous works
[11][12], but Ru-based catalysts have also been used to hydrogenate stronger C═C bonds, such as the conjugated ones in benzene rings. Regarding the IERs, both synthetic and commercial resins have been used, but mostly sulfonated and gel-type ones. Noteworthy, the anion exchange resin tested by Moreno Marrodan and coworkers yielded no significant catalytic activity in several hydrogenation reactions
[3]. Different strategies regarding the activation step of the metal catalyst emerge from the revised literature, including chemical reduction of the ionic metal with NaBH
4 and in situ reduction during reaction conditions (with no separate pre-reduction step). Depending on the author and explored reaction, the reported results differ substantially. For instance, Jeřábek reported much higher conversions when pre-reducing the catalyst at 1 MPa H
2 in water at 30 °C for 2 h compared to in situ reduction during cyclopentene hydrogenation to cyclopentane (
Table 1), whereas Moreno Marrodan reported the opposite behavior in the hydrogenation reaction of methyl 2-acetamidoacrylate to methyl acetylalaninate
[3][9].
2. Hydrogenation of Alkynes
In Drelinkiewicz et al. (2009), self-made gel-type resins functionalized by amine were used, which were obtained by suspension polymerization of a mixture of glycidyl methacrylate, ST, and diethylene glycol dimethacrylate (DEGDMA). As metal nanoparticles, Pd (0.125–0.50 wt.%) of considerably large sizes (30–50 nm) were obtained when Pd was incorporated on dry resin beads using an aqueous solution of PdCl
2 as the precursor, which contrasts with the nano-scaled Pd particles obtained when Pd(OAc)
2 was used. However, it was the 30–50 nm Pd catalyst the one exhibiting better results for the hydrogenations of 2-butyne-1,4-diol and phenylacetylene to their respective alkenes, with selectivity values ca. 92–95% at 90% conversion of the alkyne. The better performance of the large Pd nanoparticles-containing IER is related to its metal distribution (i.e., mostly, in the outer shell of the polymer beads, forming a thin layer)
[13].
3-Hexyn-1-ol was hydrogenated to 3-hexen-1-ol by Moreno Marrodan and coworkers over the already mentioned lithiated Pd/Dowex
TM 50W × 2 IER
[3]. Very high values of both conversion (98.5%) and selectivity (>99.8%) were reported in their work.
Four monometallic and six bimetallic catalysts, combining Cu, Ag, Ni, with Pd over the commercial acidic Amberlite
TM IR-120 resin in Na
+ form, were synthesized and tested for the hydrogenation of 4-nitrophenol and phenylacetylene in Silva et al. (2019). Regarding the monometallic catalysts, Pd, Ag, Ni, or Cu were ion-exchanged with the resin and subsequently reduced with NaBH
4. With regards to the bimetallic catalysts, two types are distinguished in their work: Pd-doped catalysts onto which a second metal was introduced by the same method (i.e., ion-exchange followed by NaBH
4 reduction) and M-doped catalysts (with M being either Cu, Ag or Ni) onto which Pd was introduced as the second metal
[14]. Among the monometallic catalysts, only Pd was active and selective towards ethylbenzene. Interestingly, the simultaneous presence of Ag completely inhibited the catalytic activity of Pd, while the presence of Cu enabled the selective production of styrene over ethylbenzene (i.e., >90% selectivity at high conversion values). On the other hand, Ni-containing catalysts favored ethylbenzene production. The authors attributed the superior behavior of bimetallic Cu catalysts to a geometric effect, rather than to electronic causes.
Sulman et al. (2012) and Nikoshvili et al. (2015) studied the use of hyper-crosslinked polystyrene as support for Pd nanoclusters to selectively hydrogenate C≡C bonds in acetylene alcohols (dimethylethynylcarbinol, dehydrolinalool, and dehydroisophytol)
[15][16]. The effect of the solvent nature on the observed activity was investigated, revealing that catalyst activity decreases in the order: alcohols > cyclohexane > water/ethanol mixture > octane ≥ hexane ≥ xylene > toluene > heptane, which apparently correlates to the solvent polarity. According to the authors, neither solvent–substrate interactions or hydrogen solubility in each solvent can explain the reported difference in catalytic activity and selectivity, but the strength of the solvent–catalyst interactions can successfully account for the observed activity pattern
[16].
As in the previous section, relevant alkynes hydrogenation reactions can be found in both Scheme 2 and Table 2, which also contain relevant information from the literature.
Scheme 2. Examples of hydrogenation reactions of alkenes over metal-doped IERs in the revised literature.
Table 2. Hydrogenation reactions of alkynes over metal-doped IERs in the literature.
Only gel-type IERs and hyper-crosslinked unfunctionalized polymers have been used as support for alkynes hydrogenations in the revised literature, with Pd being the most used metal. Interestingly, Ag, Ni, nor Cu showed any catalytic activity when used as monometallic catalysts supported on an Amberlite
TM resin for the hydrogenation of phenylacetylene
[14]. Again, different reduction protocols have been reported with no clear data elucidating the most favorable ones towards the hydrogenation of alkynes.
3. Hydrogenation of Carbonyl Compounds
Pd-doped, highly lipophilic gel-type IERs, based on styrene or dodecyl methacrylate, were used as catalysts for the hydrogenation of 2-ethylanthraquinone to 2-ethylanthrahydroquinone, which involves the consecutive hydrogenations of C═O and C═C bonds, in both Biffis et al. (2002) and Bombi et al. (2003). Reported selectivity for the tested ST-DVB supported catalyst was 50 and 65%, depending on the reducing step procedure (H
2 or NaBH
4, respectively), and catalytic activity was clearly lower than a benchmark industrial catalyst (Pd/silicoaluminate). On the other hand, while activity was also low, the resins based on dodecyl methacrylate reached selectivity values as high as 98%, which were slightly above the commercial alternative at the time
[17][18].
Hydrogenation of citral (3,7-dimethyl-2,6-octadienal) to geraniol (
trans-3,7-dimethyl-2,6-octadienol) and nerol (
cis-3,7-dimethyl-2,6-octadienol) was accomplished by Centomo and coworkers over Pt (0.43–2.5 wt.%) catalysts supported on two sets of functional gel-type resins synthesized by the authors, as well as a commercial 4-vinylpyridine (VP) polymer crosslinked with 2 mol% of DVB
[19]. The first set of resins contained
N,N-dimethyl-2-aminoethylmethacrylate (DMAEMA), cyanoethyl-acrylate (CEA) or methacrylic acid (MAA) as the functional monomer, and DVB as the crosslinker, while the second one contained either CEA or MAA, as the functional monomer, together with
N,
N-dimethylacrylamide (DMAA) and DVB. In addition, two different approaches to incorporate Pt into the resins were adopted: one based on the impregnation of the resin with mesitylene solutions of colloidal platinum, and another one involving immobilization of Pt precursors in pre-swollen resins, followed by chemical reduction. In addition, some of the tested resins were further modified by incorporation of a second metal, acting as Lewis acids (i.e., the introduction of 6.2–15.2 wt.% Fe
II, Co
II, or Zn
II in Pt-containing resins with –COOH groups). The authors state that improved selectivity can be achieved thanks to the presence of a more electropositive metal than the noble one because it would release electronic density to the noble active metal, which ultimately hinders C═C hydrogenation and favors that of C═O bonds. Reported conversions span from relatively low values (20–30%) to ≥80–90%, depending on the tested catalyst, with selectivity values reaching values above 90%, particularly with the bimetallic catalysts.
The C═O bond in methyl benzoylformate (methyl 2-oxo-2-phenylacetate) and 2,2,2-trifluoroacetophenone (2,2,2-trifluoro-1-phenylethan-1-one) was hydrogenated by Moreno Marrodan and coworkers over a lithiated Pd/Dowex
TM 50W × 2 IER, yielding selectivity values over 99.5% at acceptably high conversion levels (
ca. 90% and 62%) when no pre-reduction of the Pd
II species supported on the IER was conducted
[3].
Another noteworthy application of IERs in the field of carbonyl hydrogenation is that of Barbaro and associates, using Dowex
TM 50W × 2 resin as support for immobilizing an Ir complex. Their catalyst was used for the hydrogenation reactions of a number of substrates, including several imines, the α-keto ester dihydro-4,4-dimethyl-2,3-furandione, and terpenes like (
R)-carvone, with similar results to homogeneous catalysts
[20]. On the other hand, Pt-doped hyper-crosslinked polystyrene modified with cinchonidine was used as the catalyst for enantioselective hydrogenation of the C═O bond in ethylpyruvate into (
R)-ethyllactate and (
S)-ethyllactate in Bykov et al. (2009)
[21]. The authors found that enantioselectivity of activated ketone hydrogenation with platinated hyper-crosslinked polystyrene depends on the solvent, reducing agent, temperature, substrate, catalyst, and the modifier concentrations.
Examples of C═O hydrogenation reactions and relevant data are shown in Scheme 3 and Table 3.
Scheme 3. Examples of hydrogenation reactions of carbonyl groups over metal-doped IERs in the revised literature.
Table 3. Hydrogenation reactions of carbonyl groups over metal-doped IERs in the literature.
Interestingly, Pd/IER catalysts have been mostly used in the revised literature to hydrogenate C═O bonds, despite Pd being more often related to C═C hydrogenations
[12]. On the contrary, even though Ru-based catalysts are typically considered among the most favored ones for the hydrogenation of C═O bonds
[22], scarce references report on the use of Ru/IER for this type of reaction. To date, the only references available correspond to works conducted by Barbaro and coworkers (2014, 2016), which will be further addressed, since they are focused on multistep processes involving sequential C═O hydrogenation, followed by dehydration or lactonization reactions to obtain biomass-derived compounds
[23][24]. Consequently, further investigation into the potential use of Ru-doped IER for C═O hydrogenations is advised.
4. Hydrogenation of Substituted Arenes
In the cited work by Moreno Marrodan et al. (2012), hydrogenation of methyl benzoylformate and 2,2,2-trifluoroacetophenone actually constitute examples of C═O hydrogenation reactions in substituted arenes
[3]. Further on, in Moreno Marrodan et al. (2015), the number of substrates belonging to the family of substituted arenes was extended, and selective C═C hydrogenation reactions over a 1.3 wt.% Rh catalyst (3 nm) supported on Dowex
TM 50W × 2, either in lithiated or protonated form, was investigated
[25]. In their study, the effect of the bead size (i.e., beads of 38–75 µm and of 150–300 µm were used as support), the activation protocol to reduce Rh
I to Rh
0 (including no pre-reduction step, reduction with H
2, and chemical reduction with NaBH
4), and catalyst recyclability was evaluated. Results showed that as-prepared lithiated Rh-doped IERs are efficient C═C hydrogenation catalysts under undemanding conditions (room temperature and 1 bar H
2), yielding conversion and selectivity values as high as 100%, depending on the substrate.
In the already mentioned reference by Silva and collaborators
[14], besides the hydrogenation of phenylacetylene to styrene, further hydrogenation to ethylbenzene was also studied. Over 99% conversion is reported for the monometallic Pd-doped IER and the bimetallic catalysts containing Ni. Contrarily, Ag- and Cu-containing catalysts exhibited virtually nil catalytic activity. Seki and coworkers also studied the hydrogenation of benzaldehyde in supercritical CO
2 over a Pd catalyst supported on Amberlyst
TM 15
[8]. Modest results were reported in terms of both conversion (16%) and selectivity (33%).
Regarding the use of Pd-containing, hyper-crosslinked ST, the kinetics of the gas-phase phenol hydrogenation to cyclohexanone were evaluated by Sulman and coworkers
[26]. Selectivity greater than 95% is reported at 99% conversion, using Macronet™ MN270 as the Pd support.
Relevant literature examples regarding the hydrogenation of substituted arenes using IERs and related data are compiled in Scheme 4.
Scheme 4. Examples of hydrogenation reactions of substituted arenes over metal-doped IERs in the revised literature.
5. Hydrogenation of Nitroaromatic Compounds
Regarding the hydrogenation of nitroaromatics to their corresponding amines, some mentions can be retrieved from Corain et al. (2003, 2010) regarding Pd-doped gel-type IER, but with no specific data besides references contained therein
[27][28]. In Gelbard (2005), the nitrobenzene reduction to phenylhydroxylamine, with hydrazine as the reducing agent, over Pt supported on chlorinated Amberlite
TM IRA-410 is reported to achieve a 95% yield
[29].
The hydrogenation of 4-nitrophenol to 4-aminophenol was evaluated with the mentioned mono- and bimetallic catalysts supported on Amberlite
TM IR-120 resin. In general, bimetallic catalysts showed faster reaction rates than monometallic ones, except for monometallic Cu, which was the most active one. This is explained in terms of increased catalytically active sites and better availability
[14].
6. Hydrogenation of Nitrates
The aqueous catalytic reduction of nitrates to molecular nitrogen with hydrogen was studied by Gašparovičová and coworkers (2006, 2007) over bimetallic Pd-Cu catalysts (4 wt.% Pd, 1 wt.% Cu) supported on Dowex
TM 1 × 4 in the chloride form
[30][31]. Dowex
TM 1 × 4 is a gel-type ST-DVB resin functionalized with –N(CH
3)
3+ groups. In the first study, five reduction protocols were adopted (C1: NaBH
4 in ethanol; C2: NaBH
4 in water; C3: 0.1 MPa H
2, Na
2CO
3 in H
2O; C4: 0.5 MPa H
2, Na
2CO
3 in H
2O; and C5: 0.5 MPa H
2, Na
2CO
3 in methanol) and results were compared. Distinct Pd nanoparticles sizes were observed for each reduction protocol, with C1 being undetectable (which suggests either an amorphous metal phase or with very small particles), C2 presenting the largest particles (i.e., 6.7 nm), and C5 the smallest size (i.e., 2.9 nm). As for the catalytic outcome, the highest selectivity to N
2 was achieved by C1 (96%) at 71% nitrates conversion but with the largest Cu leaching detected. Overall, the best catalyst, according to the authors, was C5 since it yielded 73% conversion, with 77% selectivity towards N
2, and the lowest Cu leaching. For all catalysts, Pd leaching was below detection limits
[30].
In Gašparovičová et al. (2007), the Pd-Cu/Dowex
TM catalyst was compared to Pd-Cu/Al
2O
3. Two reduction protocols were applied to either type of catalyst. For the resin-based catalyst, the protocols were as follows: H
2 in 2.5% water solution of Na
2CO
3 for 1 h at 25 °C and 0.5 MPa (catalyst 1), or 0.066 M solution of NaBH
4 in ethanol for 1 h at ambient temperature under occasional stirring (catalyst 2). Interestingly, catalyst 2 presented a very homogeneous distribution of Pd and Cu, with Pd- and Cu-containing phases too small to be detected by XRPD, or amorphous, whereas Pd and Cu were mainly located in a shallow external layer of catalyst 1, with Pd nanoparticles of 3.8 nm. This suggests a clear effect of the reduction protocol in the metal distribution within the resin beads. Noticeably, expected byproducts, namely nitrites (produced by partial reduction) and ammonia (over reduction), were almost avoided with resin-based catalysts (i.e., selectivity towards N
2 was 93.8% at 44% conversion with catalyst 2). On the other hand, Pd-Cu/Al
2O
3 catalysts were more active, yet less selective to N
2 (below 60%) than resin catalysts. Furthermore, while Pd-Cu/Al
2O
3 catalysts showed some leaching of Cu (no leaching of Pd was detected), Cu and Pd in solution were below detection limits for resin-based catalysts
[31].
Another Pd-Cu catalyst was investigated by Mendow and coworkers for the same reaction system, with the emphasis being made on the performance of two processes for nitrates removal
[32]. In their work, Mendow used a macroporous, ST-DVB anion exchange resin (i.e., Diaion
TM WA30), containing tertiary amine functional groups. Pd incorporation in the resin was accomplished by contact with PdCl
2 solution dissolved in NaCl 0.01 M and HCl 0.01 M while bubbling N
2. After reduction of the Pd
II to Pd
0 with 35 wt.% hydrazine solution, Cu was incorporated by means of the controlled surface reduction method. In this procedure, the reduced Pd/WA30 catalyst was suspended in water and H
2 was bubbled for 2 h; then, a CuNO
3·3H
2O solution was added while maintaining the hydrogen bubbling for 2 h. After several additional steps involving filtration, reduction, and washing operations, the final metal contents were 2 wt.% Pd and 0.5 wt.% Cu, as verified by energy-dispersive X-ray fluorescence (XRF). Depending on the operating conditions, conversion, and selectivity to N
2 values as high as 100% are reported
[32].
The simultaneous catalytic reduction of nitrate ions and reductive dehalogenation of organochlorinated pollutants from water was studied by Bradu and associates, who prepared bimetallic Pd-Cu catalysts over Purolite
® A520E, which is a strong base anion exchange resin of the macroreticular type, presenting quaternary ammonium functional groups on an ST-DVB polymer matrix
[33]. Four catalysts were prepared and compared: (i) a monometallic Pd (2.01 wt.%) over A520E was prepared by ion exchange; (ii) a bimetallic Pd-Cu (1.97 wt.% and 0.45 wt.%, respectively) catalyst sample was prepared by simultaneous ion-exchange of both metal salts with the resin; (iii) a second bimetallic Pd-Cu (1.98 wt.%, 0.48 wt.%) catalyst was prepared by consecutive stages of ion-exchange of each metal salt with the resin, with Pd being exchanged in the first place; and (iv), finally, a third bimetallic Pd-Cu (2.01 wt.%, 0.49 wt.%) catalyst was prepared by Pd ion-exchange, followed by Cu deposition by controlled surface reaction. The latter catalyst sample was able to achieve selective nitrate reduction, with a 95% conversion and 92% selectivity towards N
2, with almost quantitative hydrodechlorination of 4-chlorophenol
[33].
7. One-Pot Multistep Reaction Processes Involving Hydrogenation
Regarding multistep processes, several relevant examples can be found in the literature, signaling that this specific field is of great interest since it deals directly with the integration and optimization of processes enabling the one-pot synthesis of valuable chemicals. For instance, the simultaneous hydrogenation and isomerization of diisobutylenes (2,4,4-trimethylpent-1-ene, TMP1, and 2,4,4-trimethylpent-2-ene, TMP2) over a Pd-doped IER was investigated by Talwalkar and coworkers
[34]. Interestingly, the commercial Amberlyst
TM CH28 bifunctional catalyst, which contains Pd nanoparticles and –SO
3H functional groups in a macroreticular matrix, was found to favor isomerization of TMP2 to TMP1, which is more prone to hydrogenation towards isooctane than TMP2 due to the terminal position of the double bond. From this, it follows that bifunctionalization of a resin with acid and metallic groups can offer selectivity changes in isomerization reactions, making the subsequent hydrogenation step more profitable. The effects of operating parameters (i.e., temperature, catalyst loading, hydrogen pressure, impurities associated with TMPs, and catalyst reusability) were evaluated and a kinetic model was proposed.
Seki et al. (2007, 2008) reported on the one-pot synthesis of 2-ethylhexanal from crotonaldehyde, which involves hydrogenation and aldol condensation in supercritical CO
2, over a bifunctional acidic resin-supported palladium catalyst
[8][35]. They incorporated Pd (1 wt.%) onto Amberlyst
TM 15 and obtained 13–98% conversions and ~0–67% selectivity values, depending on the assayed conditions.
The one-pot synthesis of a potential analgesic over Pd/Amberlyst
TM 15 was studied by Wissler and coworkers. In their work, two reaction steps are involved: dehydration of the starting tertiary alcohol, followed by hydrogenation of the obtained olefin. At the proposed optimized conditions, the reported conversion was 98%, and the selectivity towards the target product was 64%. However, prompt deactivation of the catalyst in the first reuse was observed, which the authors link to possible coke formation on the resin, since they argue that the selected temperature (i.e., 150 °C) is not high enough to promote thermal instability and no Pd leaching was observed
[36].
Besides the examples mentioned so far, the specific, yet broad, field of biomass transformation into high-value-added products seems to be particularly suited for using metal-doped IER. For instance, Barbaro et al. (2016) focused on the direct conversion of glucose and xylose to isosorbide and anhydroxylitol, respectively, involving hydrogenation followed by dehydration. Dowex
TM 50W × 2 in its H
+ form was used as support for Ru, with reported yields of 84.9% for isosorbide and 94.9% for anhydroxylitol at 190 °C with 30 bar H
2 in water in batch experiments
[23]. Such high yields are attributed to a combined effect of (i) a favorable microporous structure of IER and a narrow size distribution of Ru nanoparticles, which would enhance hydrogenation selectivity, and, (ii) a proper balance of density and strength of Brønsted acid sites, which would enhance dehydration.
In a different approach, a mixture of Pt/C catalyst and the thermostable, ST-DVB macroreticular resin Amberlyst
TM 70 was used for the cellulose conversion into isosorbide, which involves three consecutive reaction steps: hydration, hydrogenation, and dehydration
[37]. The same approach was adopted in Galletti et al. (2012) for the synthesis of γ-valerolactone (GVL) from levulinic acid
[38].
With respect to GVL production, Au, Pt, Ir, Ni, Cu, Re, Rh, and Ru over different supports have been mentioned as effective catalysts to obtain GVL by hydrogenation of levulinic acid
[24][39][40][41][42][43]. In this process, either hydrogenation followed by dehydration or dehydration–hydrogenation reactions take place. As for the use of IERs to produce GVL, Moreno Marrodan and Barbaro (2014) supported Ru (0.87 wt.%) on Dowex
TM 50W × 2 by contacting it with RuCl
3 and further reduction with NaBH
4, with water being used as the solvent in both metalation and reduction steps. Values of 16.2–99.8% conversions in batch runs and 89–100% conversions in continuous runs (residence time = 62–211 s, T = 70 °C, P = 4.8–7.0 bar H
2, water) are reported, with selectivity to GVL always above 99%. No significant activity decay was observed after three consecutive runs of recycling the catalyst and no Ru leaching was detected
[24].
In another study by Moreno Marrodan and coworkers, metal nanoparticles (Pd, Rh, or Ru) were supported on a perfluorinated, fluorosulfonic acid resin (i.e., Aquivion
® PFSA), which is claimed to possess an acidity similar to sulfuric acid, a thermal stability well beyond that of conventional IER, and chemical inertness in aggressive environments. Full selectivity at high conversion levels is reported for the conversion of (+)-citronellal to (-)-menthol and levulinic acid to γ-valerolactone (GVL) under mild conditions over Pd-based catalysts
[44].
Another prominent reaction system within this field is the synthesis of methyl isobutyl ketone (MIBK) from acetone, involving consecutive steps of condensation, dehydration, and hydrogenation (e.g.,
[45][46][47][48][49][50][51][52]). The commercial Pd-doped Amberlyst
TM CH28 was pinpointed as an effective catalyst to carry out this synthesis. Depending on the operating conditions, reported acetone conversion values span from 25 to >99%, with selectivity in the range 70–90%
[45][46][47][48][49][50][51]. Amberlyst
TM CH43, a Pd-doped strong acid styrenic IER, was reported to provide higher selectivity towards MIBK at high temperatures in a continuous 8 h run
[51]. Remarkably, Aquivion
® polymers were also used as Pd supports (0.10–1.04 wt.%) for the synthesis of MIBK, showing conversions in the range 11.0–63.3% and selectivity values of 62.8–96.5% in batch runs (t = 2–22 h, T = 120–180 °C, P = 10–25 bar H
2)
[52].
Several multistep reactions over metal-doped IER are shown in Scheme 5. As can be seen, all multistep processes found in the literature combine metal sites with acid sites. Both macroreticular and gel-type resins have been used. Furthermore, the commercial AmberlystTM CH28 needs to be highlighted, since it is a readily available product that has shown good results in different multistep reaction processes.
Scheme 5. Multistep reaction processes over metal-doped IERs in the revised literature.
8. Other Related Reactions
With regards to the role of IERs in other hydrogenation reactions (or related reaction systems), the direct synthesis of hydrogen peroxide from molecular oxygen and hydrogen stands out as one of the most studied processes. For instance, Sterchele and coworkers successfully prepared bimetallic Pd-Au (Pd: 1.0; Au: 0.25–1.0 wt.%) and Pd-Pt (Pd: 1.0; Pt: 0.1–1.0 wt.%) catalysts over a commercial, ST-DVB macroreticular resin, Lewatit
® K2621, by simple ion-exchange in water and reduction with aqueous formaldehyde, which they tested in the direct synthesis of hydrogen peroxide
[53]. They found that the presence of small amounts of either Pt or Au in Pd catalysts promoted selectivity while reducing activity in comparison with monometallic Pd catalysts, but that further addition of Pt or Au produced different effects: while activity increases with an increasing amount of Au, a maximum in hydrogen peroxide productivity was observed with 0.5 wt.% Pt addition, but also the lowest selectivity level was achieved. Other relevant examples of works dealing with this reaction system are those by Burato (2006) and Frison (2019), which have already been commented on in previous sections
[54][55].
Concerning catalytic hydrodechlorination reactions, Králik et al. (2014) supported Pd and Pt over Dowex
TM 1 × 4, an ST-DVB gel-type resin, functionalized with trimethylammonium, and the catalysts were tested in the hydrogenation of chloronitrobenzene isomers into the corresponding chloroanilines. The resin was used either in the chloride form or in the OH
− form, and about 1 wt.% metal contents were measured. From ISEC analyses before and after metal deposition, it became evident that lower concentrated polymer domains (i.e., 0.4 nmnm
−3) decreased with the metal incorporation while increasing higher concentrated domains (0.8 nmnm
−3)
[56]. Regarding the catalytic results,
x-chloronitrobenzene,
x-chloroaniline, aniline,
x-chloronitrosobenzene, and nitrobenzene were detected in the reaction medium after the catalytic tests, with
x-chloroaniline being the target product in that work (i.e., reduction of the NO
2 group, without hydrodechlorination). Significantly higher activity, but lower selectivity to
x-chloroaniline, was found for the Pd-containing catalysts in the OH
− form than for their chloride counterparts. As indicated by the authors, in previous works by Krátky et al. (2002, 2003), where a Pd-doped, ST-DVB sulfonated IER was compared to Pd/charcoal catalyst and the effect of different solvents was evaluated, it had already been established that condensation reactions occurring in basic environments lowered the selectivity to
x-chloroaniline
[57][58]. All Pt catalysts in Králik et al. (2014) were more active than the Pd ones, which the authors linked to the lower average size of Pt crystallites (2.3 nm) in comparison with the Pd ones (4.5 nm)
[56].
Han and coworkers used Amberlite
TM IRA-900 and IRA-958 (both in Cl
− form) as support for Pd (0.1–11.2 wt.%, 3–5 nm) to catalyze the complete hydrodechlorination of triclosan (5-chloro-2-(2,4-dichlorophenoxy)phenol). Both resins are of the macroreticular type with quaternary amine as the functional group, but IRA-900 is an ST-DVB resin whereas IRA-958 is acrylic. IRA-958 showed better performance than IRA-900, which was attributed to the hydrophobicity, pore size, and surface area of its polymeric matrix, and to larger Pd particle size (4.6 and 3.7 nm for IRA-958 and IRA-900, respectively, but with some larger aggregates of 20–40 nm diameter range observed on IRA-958). Both catalysts were reused multiple times without significant loss in effectivity or leaching
[59].
On the other hand, Jadbabaei et al. (2017) used nonionic polymeric resins (that is crosslinked Amberlite
TM XAD-4 and hyper-crosslinked Purolite
® MN200 and MN100, the latter with a small fraction of tertiary amine functional groups) and basic ST-DVB IER (i.e., Amberlite
TM IRA-910 and IRA-96, with dimethyl-ethanolammonium and tertiary amine functional groups, respectively) as supports for Pd catalysts aiming at the catalytic hydrodechlorination of 4-chlorophenol. A Langmuir-Hinshelwood kinetic model was developed, with surface reaction as the rate-determining step, which suggested an enhancing effect of adsorption on the catalytic reactivity. IRA910 was found to facilitate Cl
− adsorption and promote the production of PdCl
3− and PdCl
42− species, which are inactive for catalytic reduction
[60].
Hydrogenolysis of glycerol, which involves chemical bond cleavage with the simultaneous addition of a hydrogen atom to the resulting molecular fragments
[61], yields the production of 1,2- and 1,3-propanediol, together with ethylene glycol. Kusunoki et al. (2005), Miyazawa et al. (2006), and Centomo et al. (2013) studied this reaction system in connection to the use of IER. However, physical mixtures of carbon-supported metal catalysts (e.g., Ru/C, Pt/C, Pd/C, Rh/C) with IER were found to be much more selective towards 1,2-propanediol (which is the preferred target product) than tested metal-doped IER
[62][63][64]. Furthermore, in the framework of the hydrogenolysis of glycerol, van Ryneveld and coworkers combined Ru/C catalysts with thermal-resistant IERs, such as Amberlyst
TM DT and Amberlyst
TM 70
[65]. Regarding hydrodeoxygenation, a particular type of hydrogenolysis involving oxygen removal from oxygen-containing compounds (which is an interesting reaction system for the valorization of biomass waste streams, e.g.,
[66]), has been found to be effectively achieved by monometallic Cu/C and bimetallic Mg-Cu/C obtained from the weakly acidic, acrylic IER Diaion
TM WK11, which was calcinated after ion-exchanging it with the designated precursor salts in Wang et al. (2021, 2022). Particularly relevant features of those works are the reported high metal loadings (>50 wt.%) and small metal particle sizes (<15 nm)
[67][68].
This entry is adapted from the peer-reviewed paper 10.3390/catal13030624