Electrochemical Coproduction of Acrylate and Hydrogen from 1,3

Research Article pubs.acs.org/journal/ascecg

Electrochemical Coproduction of Acrylate and Hydrogen from 1,3Propandiol Jafar Mahmoudian,†,§ Marco Bellini,† Maria V. Pagliaro,†,§ Werner Oberhauser,† Massimo Innocenti,†,‡ Francesco Vizza,*,† and Hamish A. Miller*,† †

Istituto di Chimica dei Composti OrganoMetallici, ICCOM-CNR, Polo Scientifico Area CNR, Via Madonna del Piano 10, 50019 Sesto Fiorentino, Italy ‡ Dipartimento di Chimica, Università di Firenze, Via della Lastruccia 3, 50019 Sesto Fiorentino (FI), Italy § Università di Siena, Via Aldo Moro 2, Siena 53100, Italy S Supporting Information *

ABSTRACT: The coproduction of acrylate and hydrogen from 1,3-propandiol is achieved by electrochemical reforming using a nanoparticle Pd on mixed Vulcan XC-72 carbon and ceria electrocatalyst (Pd/C-CeO2). Electrolysis cell parameters (potential, temperature, and flow conditions) are tuned to favor the formation of acrylate with respect to the other oxidation products (3-hydroxypropanoate and malonate). Using high cell temperature combined with a low cell voltage favors the formation of acrylate, the selectivity for which is further enhanced by flowing the 1,3-propandiol solution in single-pass mode rather than recycling through the electrolyzer. Hence, at a cell temperature of 80 °C and at a fixed cell potential of 400 mV using a one pass continuous flow of 1,3-propandiol, 77% of selectivity for acrylate is obtained. KEYWORDS: Acrylate, Electrochemical reforming, Palladium catalyst, Ceria, 1,3-Propandiol

■

INTRODUCTION The fine-chemical industry is heavily dependent upon petrochemical-derived intermediates. Hence, researchers are increasingly interested in finding alternative sustainable sources of such important industrial building blocks.1,2 In particular, high value-added products can be obtained by various processes from lignocellulose biomass.3−6 However, achieving moderate yields of the desired product usually requires multiple reactor units, strong mineral acids, high temperatures and pressures, organic solvents, homogeneous catalysts, and/or difficult purification procedures. Such factors reduce the feasibility of such processes due to increased cost, waste emissions, and high energy requirements. Microbial-derived 1,3-propandiol is a convenient starting material as it is currently produced from glucose or glycerol (2013 world market volume 140 kilotons) and hence is a cheap and renewable resource.7 Acrylic acid (AA) is one of the most commonly used materials in the organic chemical industry (forecast annual production of almost 6 million tons for 2017). Current petrochemical-based AA is produced by the oxidation of propylene. Bioacrylic acid (Bio-AA) can be obtained through both chemical or biochemical routes starting from renewable compounds such as glycerol (GLY), lactic acid (LA) or C6 sugars, and ethanol.8 GLY can be oxidized in one step to AA or using an indirect pathway, where acrolein, 3-hydroxypropio© 2017 American Chemical Society

naldehyde, or acrylonitrile can be, respectively, used as intermediates.8 LA is generally dehydrated to AA using a heterogeneously catalyzed process.9 Despite the attractiveness of a biomass derived AA, no process for the transformation of GLY to AA either by the direct or the indirect pathway has been commercialized so far.8 In this work, we report the transformation of 1,3-propandiol (1,3-P) to acrylate in an electroreformer. Low temperature electrochemical processes can offer attractive solutions, as they are energy efficient, flexible, and do not require harsh conditions or multiple reagents. Electrochemical reforming or electroreforming is a prime example, as this technology combines the production of valuable chemicals from biomass-derived starting materials with the production of clean hydrogen at low temperature and atmospheric pressure.10−14 The hydrogen can be stored and transformed at will to electrical energy using a fuel cell. The electrical energy thus produced can be fed back into the electroreformer further increasing the energy efficiency of the process.13 Anion exchange membrane (AEM) alcohol electroformers operate in an alkaline environment oxidizing at the anode of an Received: March 30, 2017 Revised: May 3, 2017 Published: June 2, 2017 6090

DOI: 10.1021/acssuschemeng.7b00968 ACS Sustainable Chem. Eng. 2017, 5, 6090−6098

Research Article

ACS Sustainable Chemistry & Engineering

demonstrate that acrylate is indeed formed by the electrooxidation of 1,3-P (identified and quantified by NMR spectroscopy and HPLC). We also show that the selectivity for acrylate may be improved by operating at high temperature and low cell voltage. As anode catalyst, we use Pd nanoparticles deposited onto a mixed conductive support composed of 50 wt % Vulcan XC-72 carbon and 50 wt % CeO2. This catalyst as demonstrated in our previous studies has enhanced activity and stability for alcohol electrooxidation under alkaline conditions when compared to simple carbon-supported Pd catalysts.23,24 The CeO2 in the support contributes to a decrease in the onset electrooxidation potential, promoting the formation at low potentials of hydroxide species adsorbed on Pd (PdI−OHad), active for alcohol electrooxidation.

electrochemical cell an alcohol and evolving hydrogen at the cathode (Figure 1). Hydrogen production occurs at energy

■

EXPERIMENTAL SECTION

Carbon black (Vulcan XC-72 pellets) was purchased from Cabot Corp., USA. All metal salts and reagents were purchased from Aldrich and used as received. All aqueous solutions were freshly prepared with doubly distilled deionized water. Synthesis of Pd/C-CeO2. The anode catalyst Pd/C-CeO2 was synthesized as follows using a method described previously.25 Briefly, a mixture of Vulcan XC-72 carbon (4 g) and Ce(NO3)3·6H2O (10.1 g) in H2O (250 mL) was stirred for 60 min and then sonicated for 30 min. The pH was adjusted to 12 with KOH, and the resulting suspension was stirred for a further 2 h. The solid product was separated by filtration and washed with H2O until pH neutral. The product was dried at 65 °C and then subsequently heated under air in a tube furnace at 250 °C for 2 h. Cooling to room temperature was undertaken under a flow of Ar. The yield of C-CeO2 was 7.15 g. A portion of the C-CeO2 (4 g) was suspended in water (500 mL), stirred vigorously for 30 min, and sonicated for 20 min. To this mixture, a solution of K2PdCl4 (1.38 g) in water (60 mL) was slowly added (approximately 1 h) under vigorous stirring, followed by the addition of an aqueous solution of 2.5 M KOH (8.4 mL). Ethanol (50 mL) was added, and the resulting mixture was heated at 80 °C for 60 min. The desired product Pd/C-CeO2 was filtered off, washed several times with distilled water to neutrality, and finally dried under vacuum at 65 °C. The yield of Pd/C-CeO2 was 4.45g. Physical Characterization. X-ray Powder Diffraction (XRD). Xray powder diffraction (XRD) spectra were acquired at room temperature with a PAN analytical X’PERT PRO diffractometer, employing Cu Kα radiation (λ = 1.54187 Å) and a parabolic MPDmirror. The spectra were acquired at room temperature in the 2Θ range from 5.0 to 120.0°, using a continuous scan mode with an acquisition step size of 0.0263° and a counting time of 49.5 s. Spectra are shown in the 2Θ range of 15−90°. Electron Microscopy. Transmission electron microscopy (TEM) was performed on a Philips CM12 microscope at an accelerating voltage of 100 kV. ESI maps were recorded via the in-column omega filter spectrometer and energy-dispersive X-ray maps by an Oxford INCA Energy TEM 200 system. Samples were prepared by gently grinding the catalyst powder in agate mortars, suspending them in isopropanol, and sonicating the suspension for 20 min. The suspension was then left overnight. Some drops of the upper fraction of each suspension were finally deposited onto holey carbon-coated copper TEM grids and dried overnight. Surface Area Analysis. The BET surface area was determined using an ASAP 2020C Instrument (Micromeritics Corp.). The active metal surface area (Pd) was determined by a CO chemisorption method, adapted to carbon supported materials, at 70 °C by the use of an ASAP 2020C Instrument (Micromeritics Corp.). Before the measurements, the sample was reduced at 210 °C with H2 and treated in vacuum at same the temperature for 15 h. Cyclic Voltammetry. The electrochemical measurements were carried out using a Parstat 2273 potentiostat/galvanostat (Princeton Applied Research) equipped with a Model 616 rotating disk electrode

Figure 1. Schematic representation of the electroreformer cell.

costs of less than half when compared to state of the art water electrolyzers.11 Of equal interest is the formation of partially oxidized intermediates from the alcohol fuel. Under strongly alkaline conditions (pH ≥ 13), on palladium (Pd) electrocatalysts, complete oxidation to CO2 does not occur, as alcohols are generally partially oxidized to carboxylate compounds. From diols such as 1,2-propandiol, 1,3-propandiol, or 1,4-butandiol, the monocarboxylates (lactate, glycolate, 4hydroxy-butanoate, or acrylate) can be obtained, and these compounds are of interest as industrial intermediates in the cosmetic or polymer industries. A number of authors have recently reported their investigations regarding the transformation of alcohols into valuable chemicals in electrochemical reactors using renewable alcohols like glycerol and 1,2propandiol.10,15−17 Generally, the oxidation results in complex mixtures of partially oxidized intermediates, so a key to the effectiveness of electroreforming lies in the oxidative selectivity of the anode catalyst. The electrooxidation of 1,3-P has been studied extensively under both acidic and alkaline conditions on Pt and Au electrodes.18−22 The main products identified in alkaline media are 3-hydroxypropanoate (3-HP) and malonate (M) (Scheme 1, path a). The dehydration of 1,3-P to form acrylate through path b has been proposed by the interpretation of FTIR spectra obtained using SNIFTIR techniques.18,21 In this article, we Scheme 1. Main Electrooxidation Pathways for 1,3Propandiol under Alkaline Conditions

6091


Research Article

ACS Sustainable Chemistry & Engineering (PAR-Ametek). A 5 mm (A = 0.1963 cm2) Teflon-potted glassycarbon disk electrode tip (PINETM) was used as substrate for the deposition of the catalyst ink. The working electrode, glassy carbon (Sigradur G; 0.867 cm2), covered by the catalyst, was loaded at the top end of the cylinder. The catalyst ink consisted of Pd/C-CeO2 (22 mg) and 2-propanol (2 g). The ink was sonicated for 120 min with a Branson 3200 sonicating bath. The metal loading on each electrode was determined by weighing the amount of ink deposited on the glassy carbon disk. An Ag/AgCl electrode was used as reference, while a counter electrode consisted of a platinum wire (0.5 mm diameter). All the solutions were purged with N2 for 30 min before any electrochemical experiments. The activity of 1,3-propandiol electrooxidation was investigated in a 2 M 1,3-propandiol + 2 M KOH solution by cyclic voltammetry in the range of 0−1.20 V. Reported CVs were acquired at a scan rate of 50 mV s−1. All potentials are reported versus the reversible hydrogen electrode (RHE). Membrane Electrode Assemblies. Membrane electrode assemblies (MEAs) were prepared using the Pd/C-CeO2 anode electrocatalyst, a Fumatech anion exchange membrane, and a carbon cloth cathode containing a commercial 40 wt % Pt/C catalyst (Aldrich). The anion exchange solid polymer membrane used was fumasep FAA-3-PK-130 obtained from Fumatech mbH (Germany) which has the following characteristics: thickness (in dry form) 110−130 μm, ion exchange capacity (Cl− form) 1.1−1.4 mmol g−1, water uptake 12−25%, and ion conductivity 4.0−8.0 mS cm−1 (Cl− form). Prior to use, the membrane was converted from Br− to OH− form by soaking in 1 M KOH for 24 h followed by thorough washing in deionized water. The cathode electrode was prepared by spreading a catalyst ink onto a carbon cloth W1S1005 (CeTech Co., Ltd.) gas diffusion layer, with a Meyer rod (n°150) in order to obtain a 0.4 mg cm−2 Pt loading. The cathodic ink was prepared in a 5 mL high density polyethylene vial, mixing 200 mg of the commercial Pt (40 wt %)/C in 450 mg of distilled water, 790 mg of 1-propanol, and 1.56 g of the ionomer Nafion (5 wt % in 2propanol). The mixture was suspended with three pulses of ultrasound. The anode catalyst ink was prepared by mixing the Pd/ C-CeO2 catalyst (40 mg) with water (400 mg) and 40 mg of a 10 wt % PTFE aqueous suspension (10% PTFE final loading of the dried electrode). The resulting thick paste was spread evenly onto a 4 cm2 nickel foam support (Heze Tianyu Technology Development Co, China) to obtain a Pd loading of ca. 1 mg cm−2. The active electroreformer cell was purchased from Scribner-Associates (USA) (25 cm2 fuel cell fixture). The MEA were fabricated by mechanically pressing the anode (nickel foam), cathode Pt/C (carbon cloth), and membrane within the cell hardware.26 Electrolysis Cell Testing. The fuel, a 30 mL aqueous solution containing 2 M 1,3-propandiol (60 mmol) and 2 M KOH, was recirculated through the anode compartment at 1 mL min−1. Voltage scans and galvanostatic curves were obtained using an ARBIN BT2000 5A-4 channels potentiostat/galvanostat. Chronopotentiometry experiments were performed by applying a constant electrolysis current of 30 mA cm−2 until the cell voltage reached the value of 1.0 V. The fuel solutions after each experiment were quantitatively and qualitatively analyzed by 13C{1H} using NMR spectroscopy and HPLC. A UFLC Shimadzu chromatograph equipped with refraction index detector (RID) was used; the column is a GRACE-Alltech OA1000 organic acid (300 mm × 6.5 mm), thermostated at 65 °C. The eluent is 0.01 N H2SO4, and the eluent flow is 0.4 mL min−1. NMR spectra were acquired with a with a Bruker Avance DRX 400 spectrometer. Chemical shifts (δ) are reported in ppm relative to TMS (1H and 13C NMR spectra). Deuterated solvents (Sigma-Aldrich) used for NMR measurements were dried with activated molecular sieves; 1,4-dioxane was used as internal standard for product quantification. The hydrogen generation flow was recorded by a Bronkhorst ELFLOW mass flow meter model F-101C-002-AGD-11-V with a maximum H2 flow rate of 3 mL min−1.

reduction and deposition of 10 wt % Pd onto a mixed conductive support composed of 50 wt % carbon and 50 wt % CeO2 (C-CeO2). The XRD patterns of the C-CeO2 support and the Pd/C-CeO2 catalyst are shown in Figure 2. Most visible

Figure 2. XRD diffraction patterns of (a) 50:50 wt % C-CeO2 and (b) Pd/C-CeO2: (●) Pd, (■) CeO2, and (•) carbon.

peaks are representative of the CeO2 portion of the support while the presence of Pd metal in Pd/C-CeO2 is demonstrated by the low intensity broad peak at 40°, which indicates that the Pd nanoparticles are finely dispersed on the support with a small particle size. Chemisorption experiments show a high Pd metal surface area in the catalyst (Table 1). The average Pd particle size estimated from this data (2 nm) concurs with the mean size determined by TEM analysis (2.1 nm). Table 1. Physical Data25 sample

BET (m2 g−1)

Pd surface area (m2 g−1Pd)

Pd particle size nm (chemisorption)

Pd particle size nm (TEM)

C (Vulcan XC-72) C-CeO2 Pd/C-CeO2

222 140 145

− − 236

− − 2.1

− − 2.0

Electron Microscopy Characterization. A TEM image of a typical zone of Pd/C-CeO2 is shown in Figure 3. ESI maps of the same region help to distinguish between the carbon and CeO2 portions of the support. The ceria particles tend to form clusters and are not homogeneously distributed over the carbon support leaving portions of the carbon without contact with ceria.

■

RESULTS AND DISCUSSION Physical Characterization of Pd/C-CeO2 Anode Catalyst. The anode catalyst was prepared by the electroless

Figure 3. TEM image of a representative portion of Pd/C-CeO2 and related ESI images for C and Ce. 6092


Research Article


investigated by cyclic voltammetry (CV) (Figure 6) in both 2 M KOH and 2 M 1,3-propandiol + 2 M KOH. The CV of Pd/C-CeO2 obtained in 2 M KOH (Figure 6A) shows features characteristic for Pd-centered transitions that have been assigned by comparison to literature data as follows. On the forward anodic scan (0 to 1.4 V), there are three transitions. There is a broad peak centered at 0.4 V due to the oxidative desorption of hydrogen (A1).27 The peak at 0.6 V (A2) is ascribed to the formation of palladium hydroxides.28 At higher potentials (0.9−1.4 V), the formation of palladium oxides takes place, and this process is associated with a gradual increase in current density.29 On the reverse cathodic scan (1.4 to 0 V), the peak at 0.7 V (C1) is assigned to the reduction of Pd oxides,30 whereas the broad cathodic feature at potentials between 0.4 to 0.2 V (C2) is attributed to hydrogen adsorption.23 The electrochemical activity of Pd/C-CeO2 toward the electrooxidation of 1,3-propandiol was investigated by cyclic voltammetry in a solution of 2 M 1,3-P and 2 M KOH. The forward anodic scan shown in Figure 6B shows that the onset for 1,3-P oxidation occurs at low potentials (approximately 0.1 V) and that the oxidative current density increases continuously with increasing potential up to 1.2 V. The perturbation in the curve at around 0.8 V is associated with the formation of an amount of inactive surface palladium oxides.29 In previous studies of Pd/C-CeO2 catalysts with ethanol as substrate under the same conditions, the formation of Pd surface oxides at potentials above 0.8 V was shown to reduce almost to zero the oxidative current density.23,24 With 1,3-P, this does not occur, meaning that the presence of this polyalcohol suppresses the formation of Pd oxides in this potential range. Electroreforming Tests. The anode electrode for electroreforming testing was prepared by the application of a catalyst ink composed of Pd/C-CeO2 powder and PTFE as binder to a porous nickel foam support. The Pd metal loading was 1 mg cm−2, and the PTFE was 10 wt % of the dry catalyst layer. This anode was assembled with an anion exchange membrane (AEM) (fumasep FAA-3-PK-130)31 and a cathode electrode that consists of a commercial Pt/C (Aldrich 40 wt %) catalyst deposited on carbon cloth also with PTFE as binder (Pt loading 0.4 mg cm−2). The anode, membrane, and cathode were pressed together within the cell hardware to form the membrane electrode assembly (MEA) for testing. The fuel solution (30 mL of 2 M aqueous 1,3-propandiol and 2 M

Pd nanoparticles are visible on the carbon portions of the support as indicated in Figure 4. A statistical count (Figure 5) of these particles gives a value of 2.1 nm as the average diameter in good agreement with the results of chemisorption data (2.0 nm).

Figure 4. TEM image of Pd/C-CeO2 catalyst. White arrows indicate CeO2 clusters and black arrows individual Pd nanoparticles supported on the carbon portion of the support (scale bar is 50 nm).

Figure 5. Pd nanoparticle size distribution for Pd/C-CeO2 obtained from TEM statistical analysis.

Electrochemical Characterization. The electrochemical characteristics of the Pd/C-CeO2 anode catalyst were

Figure 6. Cyclic voltammograms of Pd/C-CeO2 in (A) 2 M KOH and (B) 2 M 1,3-propandiol + 2 M KOH electrolytes (anodic scan shown) (25 °C, 50 mV s−1). 6093


Research Article


catalyst after long periods of electrolysis is required to verify this assumption. The mean energy consumption for each experiment was determined from the integration of the instantaneous charging power over the duration of each experiment. This value, expressed in kWh per kilogram of hydrogen is reported in Table 2. The amount of hydrogen evolved during each experiment was calculated from the moles of the electrons exchanged during electrolysis. The mean value for energy consumption obtained at 60 °C (16.3 KWh/kgH2) is significantly less than those reported recently for the electroreforming of 1,3-P, with a core−shell Au−Pd/C catalyst (20.6 KWh/kgH2)33 and with an organometallic Rh-based catalyst (24 KWh/kgH2).34 The conversion values were calculated from the HPLC data (see SI for detailed description) and show that the best conversion is obtained at 60 °C (68%). Increasing the temperature to 80 °C does not improve further the conversion of 1,3-P. The main cause for incomplete conversion during batch experiments is the gradual consumption of the base in the electrolyte as the oxidation of 1,3-P progresses (one mole of KOH is consumed for each OH group oxidized). The KOH concentration drops from the initial 2 M to below 1 M. The electrolysis cell requires a highly alkaline environment to maintain the current density requested during operation. The fuel solutions at the end of each experiment were quantitatively and qualitatively analyzed by 13C{1H} NMR spectroscopy and HPLC to determine the 1,3-P oxidation products. The results are summarized in Figure 8 and show average distribution of the oxidation products obtained at 25, 60, and 80 °C (each an average of three experiments). Three oxidation products have been identified and quantified by HPLC and NMR (see Supporting Information, Figures S1 and S2). 3-Hydroxypropanoate (3-HP) was the major product detected (>50%) that results from the 4 e− oxidation of 1,3-P (Scheme 1, path a). Further oxidation in the same reaction path to malonate (M) also occurs. No formate and carbonate were detected suggesting that no C−C bond breaking occurs confirmed by carbon balance calculations (Supporting Information). In a recent FTIR study of 1,3-propanediol electrooxidation on a Pt9Bi1/C catalyst in aqueous 1 M NaOH, evidence of only two oxidation products (3-HP and M) was found.19 Interestingly, as the cell temperature increases the percentage of compounds produced from path a (3-HP and M) reduce in favor of the formation of acrylate produced through path b (8% at 25 °C and 37% at 60 °C). The presence of acrylate is confirmed by NMR spectroscopy (SI, Figure S2). The formation of acrylate in the electrochemical cell requires a combination of the electrooxidation of a single alcohol group combined with the dehydration (nonelectrochemical process) of the opposite alcohol group to form a carbon−carbon double bond (path b). Evidence of acrylate formation through the electrooxidation of 1,3-P has been suggested by an FTIR study by Lamy.18 To help understand better the reaction mechanisms, we sampled the reaction mixture periodically during a typical run over a batch performed at 80 °C (Figure 9). Initially, at low cell potentials (700 mV), the formation of acrylate slows with respect to the formation of 3-HP and M. This suggests that acrylate does not form from the sequential dehydration of 3-HP (Scheme 1, path c) but directly through path b (first dehydration of 1,3-P followed by electrooxidation). To check this, a cell was run with 2 M allyl

KOH) was recycled through the anode compartment during testing. Water was not fed to the cathode compartment as sufficient water was present for the hydrogen evolution reaction resulting from permeation through the membrane from the anode compartment. The electroreformer cell performance was first examined by cell potential scans from 0 to 1 V at 10 mV s−1 at cell temperatures of 25, 60, and 80 °C (Figure 7).

Figure 7. (A) Typical potentiodynamic scans of the electroreformer fed with an aqueous solution of 2 M 1,3-P and 2 M KOH (linear potential ramp, 10 mV s−1) (Tcell 25, 60, and 80 °C).

The electroreforming current density increases considerably with increasing temperature reaching 550 mA cm−2 at 1 V (Tcell = 80°) compared to 70 mA cm−2 at the same voltage at 25 °C. This behavior can be ascribed to the enhanced kinetics of alcohol electrooxidation with increasing temperature as has been observed in previous studies.11,32 Lamy and coworkers studied 1,3-P electrooxidation on Pt and also found an increase in current density with temperature and suggested that the process is also partially diffusion controlled.18 In order to determine the conversion efficiency of 1,3-P and identify and quantify the electrooxidation product distribution, a series of batch experiments were run by recycling a 30 mL solution of 2 M 1,3-P and 2 M KOH at the anode electrode and applying a constant electrolysis current density of 30 mA cm−2 (while monitoring the cell potential). Each batch run was stopped when the cell potential reached 1 V to avoid any contribution to the electrolysis current from water oxidation. At each of the three temperatures with the same MEA, three consecutive batches were run each time with a new fuel solution. The cell potential vs time curves are shown in Figure 8. At 25 °C, the cell potential initially is around 0.6 V and reaches 1 V after approximately 38 h. The cell voltage at 60 °C ranges from around 0.4−0.5 V at the beginning of each run reaching 1 V at around 40 h. At the highest cell temperature of 80 °C, the second and third curves differ from the first suggesting some initial instability in the system at this temperature. Additionally, no improvement in performance is gained by increasing the cell temperature from 60 to 80 °C. The conversion efficiency and energy cost (Table 2) are also slightly worse. On the other hand, in the fast potentiometric scans (Figure 7), the cell at 80 °C does perform better than at 60 °C. Under strongly alkaline conditions, at high temperatures, Pd electrocatalysts kept at high anodic potentials can gradually form inactive oxides.29 Such processes lead to longterm irreversible loss in performance. This may be the origin of the drop in performance at 80 °C. Detailed analysis of the Pd 6094


Research Article


Figure 8. Consecutive galvanostatic experiments at constant J = 30 mA cm−2 with 30 mL of fuel solution (A, 25 °C; B, 60 °C; and C, 80 °C). The bar graph shows the product distribution determined at the end of the batch experiments (average of the three experiments).

Table 2. Electroreformer Data with 1,3-Propandiol Tcell (°C)

H2 produced (mmol)

energy cost (KWh/kgH2)

conversion of 1,3-P (%)

25 60 80

90 93 81

20 16.3 18.5

63 68 57

alcohol and 2 M KOH as fuel (Tcell 60 °C). The cell produced a current density of 150 mA cm−2 at 1 V cell potential (potential scan). A galvanostatic experiment run at 60 °C (J = 30 mA cm−2) produced a conversion of 25%, and acrylate was the only product identified. Hence, acrylate formation (through path b) is favored at higher temperature and lower cell potentials. At higher cell potentials, oxidation proceeds preferentially through the direct oxidation to 3-HP (path a) without the combined dehydration reaction. By running the electroreformer at a constant potential (400 mV) rather than constant current and at 80 °C with a recycling fuel solution, the selectivity for acrylate improved to 53% (Figure 10). Further improved

Figure 9. Evolution of product distribution with increasing cell voltage (Tcell = 80 °C, J = 30 mA cm−2).

selectivity for acrylate was obtained at a constant applied potential of 400 mV at 80 °C using a single pass of the fuel 6095


Research Article


alcohol group to form 3-HP and then oxidation of the second alcohol group to form M. (2) Path b forms acrylate as the sole product. There is no evidence that acrylate forms from 1,3-HP through path c. To confirm the evolution of hydrogen during electroreforming for a typical galvanostatic experiment, we monitored the hydrogen produced using a flow meter connected to the exit of the cathode compartment. The measured H2 gas production remained stable at around 90% of the theoretical quantity calculated from the electrolysis current density (Figure S3). Finally, we have made a simple energy analysis of the electroreforming process. For the electroreformer running in batch mode at 60 °C, the cost of electrical energy input for H2 production is 16.3 kWh kgH2−1 (Table 1). The hydrogen produced has an energy content of 33 kWh kgH2−1 (LHV).37 If this H2 is fed to a state of the art PEM-FC stack with an energy efficiency of approximately 50%,38 we can obtain 16.5 kWh kgH2−1 of electrical energy. The energy output of the fuel cell matches the energy cost of the electroreformer. This can be better understood if we consider that the energy cost of producing hydrogen by electroreforming is low (about half of water electrolysis).11 The hydrogen obtained has a high energy density, of which about half can be recovered as electrical energy if used in a fuel cell. It turns out that the energy recovered from the fuel cell is about the same as the energy cost of the electroreformer. Hence, the net cost of the electroreforming will come from the chemicals consumed in the process (i.e., 1,3-P and OH−).

Figure 10. Constant cell voltage experiments at 400 mV (A) recycling fuel and (B) single pass. The bar graph shows the product distribution.

solution through the anode compartment (flow rate 0.1 mL min−1). In this way, acrylate was obtained as 77% of the oxidation products with a continuous single pass conversion of 10% (Figure 10). The current density initially dropped from 100 to 50 mA cm−2 in the first few minutes of the experiment as the 1,3-P at the electrode surface is consumed before stabilizing at around 55 mA cm−2 (controlled by the mass transport of 1,3-P from the bulk solution to the electrode surface) throughout the rest of the experiment (Figure 10) suggesting that under conditions of constant voltage and constant fuel supply to the anode the electrolysis process is stable. The selectivity for acrylate is improved using the continuous or single pass mode compared to the batch or recycling fuel solution mode. The fuel solution in single pass mode is in contact with the electrode surface for only limited time as opposed to a recycling fuel solution that is in repeated contact with the electrode. In recycling mode, the fuel composition changes considerably over the course of the experiment, which will have an influence on the oxidation mechanisms. Another important difference is the change in electrolyte concentration in the batch mode. One mole of base is consumed for every alcohol group oxidized. So, as the fuel is oxidized, the OH− ions in solution are consumed, and hence, the pH will drop. The pH affects considerably the mechanisms of alcohol electrooxidation. For example, it has been shown that for ethanol C−C bond breaking reactions occur only at pHs below 13.35 The products and intermediates of ethylene glycol oxidation on Pd are also influenced strongly by the pH.36 On the basis of the results obtained in this study, we can propose the following reaction mechanisms for 1,3-P electrooxidation under the electrochemical reforming cell conditions described in this article. There are two independent pathways, which can be favored one way or the other depending upon the imposed conditions of cell potential and temperature: (1) Path a from Scheme 1 proceeds through first oxidation of a single

■

CONCLUSIONS In this paper, we show that acrylate can be obtained by the electroreforming of 1,3-propandiol under mild conditions (25− 80 °C) and at ambient pressure. 1,3-Propandiol is currently produced industrially from biomass-derived glycerol or glucose and hence is a renewable resource. This preliminary study has demonstrated the feasibility of acrylate production using electroreforming. Electroreforming is a low energy cost technology that also produces clean and pure hydrogen that can be used to produce electrical energy in a fuel cell, thus reducing further the net energy cost of the process. In order for the process to be a practical alternative to current petrochemical-based technologies, the selectivity for the formation of acrylate compared to the other main oxidation products (3-hydroxypropanoate and malonate) must be optimized. We have shown that tuning of the process conditions (i.e., temperature, cell potential, and flow regime) can be used to achieve this. Further improvements may be obtained by tuning the anode catalyst structure, which is the subject of current work in our laboratories.

■

ASSOCIATED CONTENT

S Supporting Information *

The Supporting Information is available free of charge on the ACS Publications website at DOI: 10.1021/acssuschemeng.7b00968. Typical HPLC trace for the 1,3-P fuel solution after galvanostatic experiments, qualitative 13C{1H} NMR spectroscopic study of fuel solution after a typical galvanostatic experiment, hydrogen production rate measured during a typical galvanostatic experiment, and 6096


Research Article


■

(14) Shen, P. K.; Wang, S. L.; Hu, Z. Y.; Ll, Y. L.; Zeng, R.; Ruang, Y. Q. Hydrogen production by alcohol electrolysis. Acta Phys.-Chim. Sin. 2007, 23 (1), 107−110. (15) Chadderdon, D. J.; Xin, L.; Qi, J.; Brady, B.; Miller, J. A.; Sun, K.; Janik, M. J.; Li, W. Z. Selective Oxidation of 1,2-Propanediol in Alkaline Anion-Exchange Membrane Electrocatalytic Flow Reactors: Experimental and DFT Investigations. ACS Catal. 2015, 5 (11), 6926−6936. (16) Zhang, Z. Y.; Xin, L.; Qi, J.; Chadderdon, D. J.; Sun, K.; Warsko, K. M.; Li, W. Z. Selective electro-oxidation of glycerol to tartronate or mesoxalate on Au nanoparticle catalyst via electrode potential tuning in anion-exchange membrane electro-catalytic flow reactor. Appl. Catal., B 2014, 147, 871−878. (17) de Lucas-Consuegra, A.; Calcerrada, A. B.; de la Osa, A. R.; Valverde, J. L. Electrochemical reforming of ethylene glycol. Influence of the operation parameters, simulation and its optimization. Fuel Process. Technol. 2014, 127, 13−19. (18) Chbihi, M. E.; Takky, D.; Hahn, F.; Huser, H.; Leger, J. M.; Lamy, C. In-situ infrared reflectance spectroscopic study of propanediol electrooxidation at platinum and gold - Part 1. 1,3-propanediol. J. Electroanal. Chem. 1999, 463 (1), 63−71. (19) Gonzalez-Cobos, J.; Baranton, S.; Coutanceau, C. A Systematic in Situ Infrared Study of the Electrooxidation of C3 Alcohols on Carbon-Supported Pt and Pt-Bi Catalysts. J. Phys. Chem. C 2016, 120 (13), 7155−7164. (20) Jin, C. C.; Wang, Z. Y.; Huo, Q. S.; Dong, R. L. Different behaviors of PdAu/C catalysts in electrooxidation of propane-1,3-diol and propane-1,2-diol. Ionics 2015, 21 (3), 841−847. (21) Gomes, J. F.; Oliveira, V. L.; Pratta, P. M. P.; Tremiliosi-Filho, G. Reactivity of Alcohols with Three-Carbon Atom Chain on Pt in Acidic Medium. Electrocatalysis 2015, 6 (1), 7−19. (22) Dimos, M. M.; Blanchard, G. J. Examining the Electrocatalytic Oxidation of Selected Diols at Nanoporous and Planar Pt Electrodes. J. Phys. Chem. C 2011, 115 (22), 11247−11256. (23) Bambagioni, V.; Bianchini, C.; Chen, Y. X.; Filippi, J.; Fornasiero, P.; Innocenti, M.; Lavacchi, A.; Marchionni, A.; Oberhauser, W.; Vizza, F. Energy Efficiency Enhancement of Ethanol Electrooxidation on Pd-CeO2/C in Passive and Active Polymer Electrolyte-Membrane Fuel Cells. ChemSusChem 2012, 5 (7), 1266− 1273. (24) Wang, L. Q.; Lavacchi, A.; Bevilacqua, M.; Bellini, M.; Fornasiero, P.; Filippi, J.; Innocenti, M.; Marchionni, A.; Miller, H. A.; Vizza, F. Energy Efficiency of Alkaline Direct Ethanol Fuel Cells Employing Nanostructured Palladium Electrocatalysts. ChemCatChem 2015, 7 (14), 2214−2221. (25) Miller, H. A.; Lavacchi, A.; Vizza, F.; Marelli, M.; Di Benedetto, F.; D'Acapito, F. D. I.; Paska, Y.; Page, M.; Dekel, D. R. A Pd/C-CeO2 Anode Catalyst for High-Performance Platinum-Free Anion Exchange Membrane Fuel Cells. Angew. Chem., Int. Ed. 2016, 55 (20), 6004− 6007. (26) Miller, H. A.; Bellini, M.; Vizza, F.; Hasenohrl, C.; Tilley, R. D. Carbon supported Au-Pd core-shell nanoparticles for hydrogen production by alcohol electroreforming. Catal. Sci. Technol. 2016, 6, 6870. (27) Pattabiraman, R. Electrochemical investigations on carbon supported palladium catalysts. Appl. Catal., A 1997, 153 (1−2), 9−20. (28) Hu, C. C.; Wen, T. C. Voltammetric Investigation of Palladium Oxides 0.1. Their Formation Reduction in Naoh. Electrochim. Acta 1995, 40 (4), 495−503. (29) Wang, L. Q.; Lavacchi, A.; Bellini, M.; D’Acapito, F.; Di Benedetto, F.; Innocenti, M.; Miller, H. A.; Montegrossi, G.; Zafferoni, C.; Vizza, F. Deactivation of Palladium Electrocatalysts for Alcohols Oxidation in Basic Electrolytes. Electrochim. Acta 2015, 177, 100−106. (30) Grden, M.; Lukaszewski, M.; Jerkiewicz, G.; Czerwinski, A. Electrochemical behaviour of palladium electrode: Oxidation, electrodissolution and ionic adsorption. Electrochim. Acta 2008, 53 (26), 7583−7598. (31) Carmo, M.; Doubek, G.; Sekol, R. C.; Linardi, M.; Taylor, A. D. Development and electrochemical studies of membrane electrode

details of calculations of conversion, selectivity, and carbon balance. (PDF)

AUTHOR INFORMATION

Corresponding Authors

*E-mail: [email protected] (F. Vizza). *E-mail: [email protected] (H. A. Miller). ORCID

Hamish A. Miller: 0000-0003-1668-6476 Author Contributions

The manuscript was written through contributions of all authors. All authors have given approval to the final version of the manuscript. Notes

The authors declare no competing financial interest.

■

ACKNOWLEDGMENTS We acknowledge the Ente Cassa di Risparmio di Firenze (project EnergyLab).

■

REFERENCES

(1) Sheldon, R. A. Engineering a more sustainable world through catalysis and green chemistry. J. R. Soc., Interface 2016, 13 (116), 20160087. (2) Sheldon, R. A. Green and sustainable manufacture of chemicals from biomass: state of the art. Green Chem. 2014, 16 (3), 950−963. (3) Yung, M. M. Catalytic Conversion of Biomass to Fuels and Chemicals. Top. Catal. 2016, 59 (1), 1−1. (4) Lin, H. F.; Biddinger, E. J.; Mukarakate, C.; Nimlos, M.; Liu, H. C. Transformations of Biomass and its derivatives to Fuels and Chemicals Preface. Catal. Today 2016, 269, 1−1. (5) Sheldon, R. A.; Sanders, J. P. M. Toward concise metrics for the production of chemicals from renewable biomass. Catal. Today 2015, 239, 3−6. (6) FitzPatrick, M.; Champagne, P.; Cunningham, M. F.; Whitney, R. A. A biorefinery processing perspective: Treatment of lignocellulosic materials for the production of value-added products. Bioresour. Technol. 2010, 101 (23), 8915−8922. (7) Saxena, R. K.; Anand, P.; Saran, S.; Isar, J. Microbial production of 1,3-propanediol: Recent developments and emerging opportunities. Biotechnol. Adv. 2009, 27 (6), 895−913. (8) Katryniok, B.; Bonnotte, T.; Dumeignil, F.; Paul, S. Production of Bioacrylic Acid. In Chemicals and Fuels from Bio-Based Building Blocks; Wiley-VCH Verlag GmbH & Co. KGaA, 2016; pp 217−244. (9) Xu, X. B.; Lin, J. P.; Cen, P. L. Advances in the research and development of acrylic acid production from biomass. Chin. J. Chem. Eng. 2006, 14 (4), 419−427. (10) de la Osa, A. R.; Calcerrada, A. B.; Valverde, J. L.; Baranova, E. A.; de Lucas-Consuegra, A. Electrochemical reforming of alcohols on nanostructured platinum-tin catalyst-electrodes. Appl. Catal., B 2015, 179, 276−284. (11) Chen, Y. X.; Lavacchi, A.; Miller, H. A.; Bevilacqua, M.; Filippi, J.; Innocenti, M.; Marchionni, A.; Oberhauser, W.; Wang, L.; Vizza, F. Nanotechnology makes biomass electrolysis more energy efficient than water electrolysis. Nat. Commun. 2014, 5, na DOI: 10.1038/ ncomms5036. (12) Caravaca, A.; de Lucas-Consuegra, A.; Calcerrada, A. B.; Lobato, J.; Valverde, J. L.; Dorado, F. From biomass to pure hydrogen: Electrochemical reforming of bio-ethanol in a PEM electrolyser. Appl. Catal., B 2013, 134-135, 302−309. (13) Bambagioni, V.; Bevilacqua, M.; Bianchini, C.; Filippi, J.; Lavacchi, A.; Marchionni, A.; Vizza, F.; Shen, P. K. Self-Sustainable Production of Hydrogen, Chemicals, and Energy from Renewable Alcohols by Electrocatalysis. ChemSusChem 2010, 3 (7), 851−855. 6097


Research Article

ACS Sustainable Chemistry & Engineering assemblies for polymer electrolyte alkaline fuel cells using FAA membrane and ionomer. J. Power Sources 2013, 230, 169−175. (32) Chen, Y.; Bellini, M.; Bevilacqua, M.; Fornasiero, P.; Lavacchi, A.; Miller, H. A.; Wang, L.; Vizza, F. Direct alcohol fuel cells: toward the power densities of hydrogen-fed proton exchange membrane fuel cells. ChemSusChem 2015, 8 (3), 524−33. (33) Miller, H. A.; Bellini, M.; Vizza, F.; Hasenohrl, C.; Tilley, R. D. Carbon supported Au-Pd core-shell nanoparticles for hydrogen production by alcohol electroreforming. Catal. Sci. Technol. 2016, 6 (18), 6870−6878. (34) Bellini, M.; Filippi, J.; Miller, H. A.; Oberhauser, W.; Vizza, F.; He, Q.; Grützmacher, H. Hydrogen and Chemicals from Renewable Alcohols by Organometallic Electroreforming. ChemCatChem 2017, 9 (5), 746−750. (35) Fang, X.; Wang, L. Q.; Shen, P. K.; Cui, G. F.; Bianchini, C. An in situ Fourier transform infrared spectroelectrochemical study on ethanol electrooxidation on Pd in alkaline solution. J. Power Sources 2010, 195 (5), 1375−1378. (36) Wang, L. Q.; Meng, H.; Shen, P. K.; Bianchini, C.; Vizza, F.; Wei, Z. D. In situ FTIR spectroelectrochemical study on the mechanism of ethylene glycol electrocatalytic oxidation at a Pd electrode. Phys. Chem. Chem. Phys. 2011, 13 (7), 2667−2673. (37) Midilli, A.; Ay, M.; Dincer, I.; Rosen, M. A. On hydrogen and hydrogen energy strategies I: current status and needs. Renewable Sustainable Energy Rev. 2005, 9 (3), 255−271. (38) Pellow, M. A.; Emmott, C. J. M.; Barnhart, C. J.; Benson, S. M. Hydrogen or batteries for grid storage? A net energy analysis. Energy Environ. Sci. 2015, 8 (7), 1938−1952.

6098


Electrochemical Coproduction of Acrylate and Hydrogen from 1,3

Recommend Documents