Surfactant-Free Synthesis of Carbon-Supported Palladium

Subscriber access provided by Olson Library | Northern Michigan University

Article

Surfactant-Free Synthesis of Carbon Supported Palladium Nanoparticles and Size Dependent Hydrogen Production from Formic Acid-Formate Solution Shuo Zhang, Bei Jiang, Kun Jiang, and Wen-Bin Cai ACS Appl. Mater. Interfaces, Just Accepted Manuscript • DOI: 10.1021/acsami.7b08441 • Publication Date (Web): 28 Jun 2017 Downloaded from http://pubs.acs.org on June 30, 2017

Just Accepted “Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted online prior to technical editing, formatting for publication and author proofing. The American Chemical Society provides “Just Accepted” as a free service to the research community to expedite the dissemination of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts appear in full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been fully peer reviewed, but should not be considered the official version of record. They are accessible to all readers and citable by the Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered to authors. Therefore, the “Just Accepted” Web site may not include all articles that will be published in the journal. After a manuscript is technically edited and formatted, it will be removed from the “Just Accepted” Web site and published as an ASAP article. Note that technical editing may introduce minor changes to the manuscript text and/or graphics which could affect content, and all legal disclaimers and ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors or consequences arising from the use of information contained in these “Just Accepted” manuscripts.

ACS Applied Materials & Interfaces is published by the American Chemical Society. 1155 Sixteenth Street N.W., Washington, DC 20036 Published by American Chemical Society. Copyright © American Chemical Society. However, no copyright claim is made to original U.S. Government works, or works produced by employees of any Commonwealth realm Crown government in the course of their duties.

Page 1 of 29

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

ACS Applied Materials & Interfaces

Surfactant-Free Synthesis of Carbon Supported Palladium Nanoparticles and Size Dependent Hydrogen Production from Formic Acid-Formate Solution Shuo Zhang†, Bei Jiang†, Kun Jiang‡*, Wen-Bin Cai†*

† Shanghai Key Laboratory of Molecular Catalysis and Innovative Materials, Collaborative Innovation Center of Chemistry for Energy Materials, Department of Chemistry, Fudan University, Shanghai 200433, China. E-mail: [email protected] ‡ Rowland Institute, Harvard University, Cambridge, MA 02142, United States. E-mail: [email protected]

Abstract Steerable hydrogen generation from the hydrogen storage chemical formic acid via heterogeneous catalysis has attracted considerable interest given the safety and efficiency concern in handling H2. Herein, a series of carbon supported capping agent-free Pd nanoparticles (NPs) with mean sizes tunable from 2.0 to 5.2 nm are developed on the demand of more efficient dehydrogenation from a formic acid-formate solution of pH 3.5 at room temperature. The trick for the facile size-controlled synthesis of Pd/C catalysts is the selective addition of Na2CO3, NH3·H2O or NaOH to a Pd(II) solution to attain initial pH values of 7 to 9.5. For comparison, cuboctahedron modeling and electrochemical COads stripping methods are applied to evaluate active surface Pd sites for turnover frequency (TOF) calculation. Both mass activity and specific activity (TOF) of hydrogen production are not only time dependent but also Pd-size dependent. Initial H2 production rate of 246 L·h-1·g-1Pd is achieved on 2.0-nm Pd/C at 303 K, together with a TOF of 1815 h-1 on the basis of cuboctahedron modeling of surface active Pd sites. The initial TOF exhibits a significant rise from 3.5 down to 2.8 nm, then levels off below 2.8 nm and even shows a maxima at ca. 2.2 nm using the electrochemical surface area for calculation. The volcano-shaped dependence of TOF on Pd NP size may be better attributed to the changing ratios of terrace sites to defect sites on Pd NPs.

Keywords: hydrogen production, dehydrogenation of formic acid, palladium catalyst, size controlled synthesis, size effect

1

ACS Paragon Plus Environment


1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 2 of 29

Table of Contents Graphic

2


Page 3 of 29

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


Introduction Hydrogen (H2) is widely regarded as a highly efficient clean energy carrier.1-2 However, the storage and transportation problem has restricted its practical application.3-4 Much effort was devoted to the exploration of new ways of hydrogen storage and production including using metal hydrides5, nitrogen/boron compounds6 and MOF7. Formic acid, which has a hydrogen mass capacity of 4.4 wt.%, is recently considered as one of the promising chemical materials for hydrogen storage for its enormous availability and convenient transport.8-11 Formic acid decomposition (FAD) may proceed according to the following two pathways: Dehydrogenation: HCOOH → H2 + CO2 Dehydration: HCOOH → CO + H2O The dehydrogenation pathway is highly demanded in FAD for the hydrogen production purpose while the dehydration pathway should be avoided with best effort. The reactivity and selectivity of the two pathways strongly rely on catalysts, and varieties of heterogeneous and homogeneous catalysts have been developed for FAD.12-22 Compared to homogeneous catalysis, heterogeneous catalysis takes the advantages of low temperature operation and easy separation. Among the most widely studied Au,23-24 Pt25-26 and Pd27-28 based heterogeneous catalysts, the Pd-based ones are the most attractive given the fact that Pd possesses much lower cost and higher activity towards FAD as compared to Au and Pt.29 A number of efforts have been devoted to achieve higher FAD performances on Pd-based catalysts, including chemical and structural modification either on Pd nanoparticles or supports.30 On one hand, Pd nanoalloys (such as PdAg31-32, PdAu33-34, PdK35, PdNi36, PdNiAg37, PdNiAu38) and core-shell structures (Ag@Pd39, Au@Pd40, AgPd@Pd41 and PdAu@Au42) were introduced to improve the Pd mass activity. On the other hand, new carbon materials (undoped and doped reduced graphene oxide 43-46 and nanoporous carbon47), silica48, titanium dioxide41, zeolite49, basic resin50 and metal-organic framework51-52 were also examined for the supports. One of the major concerns in those reports was the controlled size and dispersion of Pd-based nanoparticles. Unfortunately, the complication in synthesis and post-treatment of these new supports and Pd-based nanoparticles may severely limit their wide applications. Specifically, the organic surfactants were commonly used to synthesize Pd nanoparticles with smaller sizes,49, 53-56 as a result, extra effort should be spared to remove them in order to avoid the activity loss. A promising solution to address this cost-effective issue is the facile synthesis of ultrafine Pd NPs supported on the most common carbon black in aqueous solution without adding any organic surfactants. Furthermore, it is well recognized that catalytic and electrocatalytic performances can vary with Pd NP sizes.57-60 However, it was still quite controversial regarding the size dependent Pd-catalyzed FAD. Chan et al suggested that the optimal particle size may be located in the range of 1.8-3.5 nm without providing any experimental evidences.61 Yamashita et al examined the catalytic activity of FAD on Pd NPs in the size range of 2.7 to 5.5 nm in a formic acid solution, and indicated that the mean Pd nanoparticle size 3.9 nm afforded the best catalytic performance.62 Au-Pd alloy NPs with sizes from 3.8 to 12.8 nm on carbon black were synthesized by using different amino acids as structure-directing agents, and the 3.8-nm AuPd/C gave the highest initial TOF value, or 718 h-1 for hydrogen production in formic acid-formate solution at room 3



1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 4 of 29

temperature.63 It should be noted that, in the early exploration of size dependent catalytic behavior, organic capping agents such as PVP and/or CTAB that were employed to control Pd NPs sizes may lead to uncertainty for the surface-sensitive catalysis. Therefore, to better clarify the size effect, it is necessary albeit challenging to scale down capping agent-free Pd NPs on carbon black and revisit their size-dependent activities toward hydrogen production without introducing additional surfactant or compositional effect. In this work, a facile surfactant-free impregnation method is developed to prepare a series of carbon supported Pd NPs with a mean Pd NP size tunable from 2.0 to 5.2 nm through controlling the pH values and selecting an alkali in the initial Pd(II) precursor. The room-temperature hydrogen production is explored on these Pd/C catalysts with varying Pd NP sizes in a formic acid-sodium formate (FA-SF) solution of pH3.5. Pd size effects on mass and specific activities toward hydrogen production are discussed in details. Cuboctahedron modelling and anodic CO stripping methods are comparatively applied to evaluate the surface active sites used for calculating TOF values.

Experimental Chemicals Palladium chloride dihydrate and carbon black (Vulcan XC-72) were purchased from Aldrich and Cabot Co., respectively. PdCl2 was dissolved in 0.1 M NaCl to form 0.05 M Na2PdCl4. Sodium borohydride (A.R., purity ≥ 96%) was purchased from Sinopharm Chemical Reagent, and sodium carbonate anhydrous (AR., purity ≥ 99.8%) from Chinasun Specialty Products. All chemicals were used as received without further purification. Preparation of catalysts Pd/C catalysts with a Pd loading of 5 wt.% were synthesized by an impregnation-reduction method using borohydride (NaBH4) as the reducing agent. Typically, 318 mg of Vulcan XC-72 and 3.15 ml of 50 mM Na2PdCl4 were added into 40 mL of Milli-Q water (18.2 MΩ·cm). The mixture was sonicated for 20 min and kept under vigorous stirring for another 4 h. Then, the solution pH was adjusted by selective addition of a certain amount of Na2CO3, NaOH or NH3H2O as indicated in Table 1. Then, 10 mL of 0.05 M Na2CO3 containing 30 mg of NaBH4 was added dropwise to the suspension with a vigorous stirring by a peristaltic pump at 0.5 mL/min.64 The suspension was stirred continuously for 12 h. At last, the suspension was filtered and washed with copious amounts of Milli-Q water and then dried in vacuum at 343 K overnight. To ensure the reproducibility of the synthesis, the synthesis temperature was controlled at 290 K precisely. Characterization of catalysts The metal loading of a Pd/C catalyst was analyzed by inductively coupled plasma-atomic emission spectroscopy (ICP-AES) on a Hitachi P-4010 spectrometer. Prior to the ICP-AES measurement, the Pd component of the catalyst sample was dissolved in a hot aqua regia. The morphology and size distributions of Pd nanoparticles were characterized by high resolution transmission electron microscopy (HR-TEM) using JEOL JEM-2010 microscope. Surface species on Pd NPs were analyzed by X-ray photoelectron spectroscopy (XPS) with a Perkin-Elmer PHI-5000C ECSA system equipped with a hemispherical electron energy analyzer using a 4


Page 5 of 29

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


monochromatic Mg Kα-radiation (1253.6 eV). The binding energies were calibrated with reference to the C 1s peak at 284.6 eV. Electrochemical measurement The electrochemical measurement was performed on a CHI 605B electrochemistry workstation. The working electrode was prepared as follows: 2.0 mg of Pd/C, 1.0 mL of C2H5OH and 120 µL of Nafion was ultrasonically mixed into a catalyst ink, then 10 µL of the ink was dropped onto a freshly polished glassy carbon electrode. A Au sheet and a Hg/Hg2SO4 electrode were used as the counter and reference electrodes, respectively. CO stripping voltammograms began with bubbling CO (>99.9% purity) over the working electrode at -0.6 V for 15 min. Then the dissolved CO was removed from the electrolyte by bubbling N2 for 55 min with the electrode potential controlled at 0.0 V. In the end, the CO stripping voltammograms were recorded between -0.45 and 0.65 V at a scan rate of 10 mV s-1. Hydrogen production measurement Typically, 100 mg of 5 wt.% Pd/C was placed in a flask, then 5 mL of the mixture solution containing 5.5 mmol of formic acid and 4 mmol of sodium formate (i.e., 1.1 M FA + 0.8 M SF) was quickly injected into the flask under a magnetic stirring control. The reaction temperature was kept at 303 K and the stirring rate at 750 rpm. The produced gases were analyzed on a gas chromatograph GC 2060 (Ramiin Company) with a thermal conductivity detector (TCD) and a hydrogen flame ionization detector (FID) whose detection limit for CO is 1 ppm. An Ar balanced gas mixture containing 5.01 vol.% CO2, 13.30 vol.% H2, 1.02 vol.% CO and some other alkanes were used as the reference gas. The UV-vis spectra were recorded on a PE-Lambda 35.

Results and discussion Physicochemical Characterization Pd/C catalysts were characterized by using ICP-AES, TEM and XPS. The ICP-AES result indicated that the Pd loading was 4.9 ± 0.1wt.% for all the catalysts, nearly identical to the preset loading 5.0 wt.%. Figure 1 shows typical TEM images of different Pd/C samples and the corresponding distribution histograms of Pd particle-sizes. The Pd NPs disperse rather homogenously on carbon support without obvious aggregation. The histograms of particle-size distribution by statistically counting more than 200 nanoparticles from TEM images for different areas of each sample are shown in Figure 1. The particle size distribution gradually broadens with increasing mean size. By adding a pH-adjusting reagent selectively from three candidates Na2CO3, NaOH and NH3·H2O, the average Pd particle sizes can be flexibly controlled in the range of 2.0 to 5.2 nm. Table 1 lists the as-synthesized Pd/C samples with different Pd sizes together with the corresponding preparation conditions. Notably, the precursor solution containing 4 mM Na2CO3 yields the smallest mean size 2.0 nm for Pd NPs.

5



1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 6 of 29

Figure 1. Representative (high resolution) TEM images and corresponding histograms of Pd particle-size distributions of as-synthesized Pd/C catalysts: (a) 2.0 nm, (b) 2.2 nm, (c) 2.5 nm, (d) 2.8 nm, (e) 3.6nm, (f) 5.2 nm. Scale bars are 20 nm and 2 nm for lower magnitude and high-resolution images, respectively.

Table 1.

Synthetic conditions and corresponding Pd particle sizes

Additive

Concentration (mM)

Initial pH value of precursor solution

Particle size (nm)

Na2CO3

2

7.0

2.5±0.4

Na2CO3

4

7.5

2.0±0.3

Na2CO3

8

8.6

2.2±0.4

Na2CO3

20

9.5

2.8±0.6

NH3· H2O

10

7.5

3.6±0.6

NaOH

0.04

7.5

5.2±1.4

The initial pH values and the complexing states in the precursor solutions significantly affect the Pd NP size. 54, 63 In the previous reports based on NaBH4 reduction, the pH values of precursor solutions were usually controlled to around 9 by adding NaOH or ammonium hydroxide, resulting in a mean particle size of ca. 3 nm. It is very difficult if not impossible to attain particle mean size down to 2.0 nm without using organic or polymeric surfactants such as PDA and PVP. To our best knowledge, Pd NPs down to 2.0 nm have been for the first time obtained herein by simply adding appropriate amount of Na2CO3 in the precursor solution without introducing the above-mentioned surfactants. UV-vis spectroscopy was used to characterize the structural difference of Pd(II)-containing complexes in the precursor solutions upon addition of different pH-adjusting reagents in the absence of carbon black. Representative UV-vis spectra are shown in Figure 2. The [PdCl3(H2O)]complex may be present in the aqueous solution of Na2PdCl4 before adding a pH-adjusting reagent.65 6


Page 7 of 29

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


Figure 2. UV-vis spectra of aqueous precursor solutions It is well known that the coordination of metal ions changes the redox potential of Mn+/M and initial pH of a precursor solution also affect the reducing strength of NaBH4. Thus the metal nucleation should be dependent on the initial coordination structures of the complex metal ions as well as the initial pH with dropwise addition of the same reducing agent NaBH4. With 4 mM Na2CO3 in the precursor solution, the 245-nm absorption band is greatly attenuated with a new band at 260 nm showing up, corresponding to the conversion from [PdCl3(H2O)]- to [PdClx(OH)4-x]2- as a result of the right shifted hydrolysis reaction CO32- + H2O ↔ HCO3- + OH-. 47, 65 In contrast, the UV-vis spectra for the 4 mM Pd(II) solutions in the absence and presence of 0.04 mM NaOH are quite close, indicating a negligible formation of [PdClx(OH)4-x]2- species. Nevertheless, according to a previous report,47 further increasing NaOH concentration led to the 260-nm peak exactly same as that in Figure 2 with the addition of Na2CO3, in support of the assignment of Pd[Clx(OH)4-x]2- for that peak. It also can be seen from Figure 2 that the 260-nm band was essentially stabilized when the Na2CO3 concentration reaches 4 mM, at which the ligand exchange is likely completed. On the other hand, with the introduction of 10 mM NH3·H2O, the original [PdCl3(H2O)]- complex could be partially converted to [PdClx(NH3)4-x]2-x, as evidenced by an attenuated UV absorption below 260 nm and an enhanced absorption at longer wavelengths, compared with the absorption spectrum of 4 mM Na2PdCl4 solution. Table 1 indeed confirms that the initial pH and the complexing state of Pd(II) in a precursor solution coaffect the final particle size, although full understanding of these two coeffects is quite difficult at this moment and beyond our focus in the present work. The surface compositions and valences of different-sized Pd nanoparticle supported on carbon black are examined by ex situ X-ray photoelectron spectroscopy (XPS) shown in Figure 3a. By deconvoluting the Pd 3d5/2 and 3d3/2 peaks, the coexistence of Pd(0) and Pd(II) species can be confirmed, and the percentages of Pd(0) and Pd(II) species for different Pd/C catalysts are shown in the bar diagram of Figure 3b. The fractional ratio of Pd(II) to Pd(0) species is increased with decreasing Pd particles, suggesting that more surface active area exposed in ambient condition on 7



1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 8 of 29

smaller particles, in contrast to a previous report that this ratio was independent of the Pd particle size.62 The present result is in accordance with the very good dispersion of Pd NPs on carbon black, in considering that for a given mass of well-dispersed Pd NPs, a smaller particle size comes with a larger surface area. Also noted is that the 3d5/2 and 3d3/2 peak positions for Pd(0) are largely independent of Pd particle size, in other words, they share nearly same electronic states for all Pd/C catalysts. This may indicate an insignificant size dependent electronic effect over the range of 2.0 to 5.2 nm for Pd/C catalyzed FAD. It may be necessary to point out that these partially oxidized Pd surfaces in the ambient atmosphere can be readily reduced during the hydrogen production from FA-SF electrolyte where the local potential of Pd nanoparticles is close to 0 V RHE.

Figure 3. (a) Core level XPS spectra for Pd3d regions and (b) percentages of Pd(0)and Pd(II) species as determined from the deconvoluted Pd3d doublet peaks of Pd/C catalysts with different sizes. Catalytic performance To further examine the size dependent FA dehydrogenation activity on Pd surfaces, we have measured the gas produced as a function of time on the above Pd/C catalysts in a mixed solution of 1.1 M formic acid and 0.8 M sodium formate. Sodium formate was proved to play an important role in accelerating dehydrogenation by inducing a favorable adsorption orientation of formic acid (H-down configuration) or serve as the reactive intermediate in the process of FAD.66 pH close to pKa favors mostly the dehydrogenation rate.67-68 The released gaseous mixture batches were collected in gas burettes and the componential concentrations were analyzed by gas chromatography (GC). Gas chromatograms are shown in Figure 4. The GC analysis confirmed no CO was produced over the whole FAD process, and the volumetric ratio of H2 to CO2 is around 1.05:1 possibly owing to a little dissolution of CO2 into the reactant system in the test process.

8


Page 9 of 29

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


Figure 4. Gas chromatograms obtained using (a) TCD and (b) FID for the FAD gas from the FA-SF aqueous solution in the presence of 100 mg of 2.2-nm Pd/C catalyst.

Figure 5. (a) Time course of gas evolution from 5 mL of 1.1 M FA + 0.8 M SF solution in the presence of 100 mg of Pd/C catalysts (ca. 5 mg of Pd) with different Pd particle sizes. (b)Temperature dependence of gas evolution from the aqueous FA-SF solution in the presence of 2.2-nm Pd/C catalyst. The inset is Arrhenius plot of ln(TOF) vs 1/T. The plots of hydrogen production mass activities (c) and TOF values (d) versus Pd NP size for 2 min, 5 min and 8 min of FAD. Figure 5(a) shows the hydrogen volume versus time curves measured respectively for Pd/C catalysts of different Pd particle sizes in the mixed FA-SF solutions. The FAD rates are flattened out within 15 min for Pd NP sizes smaller than 3 nm. Nearly 100% of HCOOH species decomposes to H2 and CO2 with Pd sizes from 2.0 to 2.8 nm, and ca. 90% and 78% of HCOOH species do so with 3.6-nm and 5.2-nm Pd/C catalysts, respectively. Turnover frequency (TOF) 9



1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 10 of 29

values for the FAD reaction are used to evaluate the catalytic performances of the Pd/C catalysts. Here TOF is defined as

TOF=

As the equilibrium shapes of fcc metals are cuboctahedra, the cuboctahedron model has been widely used to estimate the surface areas of Pd and Pt nanoparticles in literatures.69-70 Along this lien, the modelled cuboctahedra made up of concentric shells of atoms were adopted herewith to simulate the actual Pd nanoparticles. The number of surface atoms may be calculated as follows, 71-72

=

/()(" ")

= N .

m ≈ d / 0.45 where m is the total number of atomic layers or cluster edge length and N is the numeric ratio of surface atoms to total atoms. Table 2. Comparison of H2 generation rates and TOF values for differently sized Pd/C catalysts in contact with 5 mL of 1.1 M FA + 0.8 M SF solution for 2, 5 and 8 min at 303 K. Cuboctahedron model is used for counting surface Pd sites. 2 min Pd particle size (nm)

N

H2 generation rate -1

-1

5 min TOF -1

H2 generation rate -1

-1

8 min TOF -1

H2 generation rate

TOF

-1

-1

(L h gPd )

(h )

(L h gPd )

(h )

(L h gPd )

(h-1)

2.0±0.3

0.58

246

1815

198

1461

172.5

1272

2.2±0.4

0.53

216

1744

177.6

1434

159

1283

2.5±0.4

0.48

189

1685

154.8

1380

139.5

1243

2.8±0.6

0.44

165

1604

136.8

1330

124.5

1211

3.6±0.6

0.35

69

843

58.8

719

57

697

5.2±1.4

0.25

48

821

38.4

657

33

565

The surfactant-free Pd/C catalysts with smaller Pd sizes showed superb performance toward catalytic hydrogen production. The 2.0-nm Pd/C yielded an initial TOF of 1815 h-1 and a H2 production rate of 246 L h-1gPd-1 over the first 2 min (corresponding to ca. 30% FA consumed), surpassing most heterogeneous and homogeneous catalysts in similar catalytic environments.43, 47, 63, 73

Temperature is known to significantly affect the FAD course. The H2 production volume as a function of time was measured at different temperatures with the 2.2-nm Pd/C in 1.1 M FA + 0.8 M SF (Figure 5b). The TOF values over initial 2 min are 1744, 2809, 3850 and 5521 h-1, corresponding to 303, 313, 323 and 333 K, respectively. According to Arrhenius equation, the plot of ln(TOF) vs 1/T gives an apparent activation energy (Ea) of 31.7 KJ mol-1, which is among the smallest ones reported for heterogeneous and homogeneous FAD.61, 74 10


Page 11 of 29

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


Figure 6. Hydrogen production volume as a function of time over the 2.2-nm Pd/C catalyst in FA-SF solution at 303 K for three consecutive runs. The cycle stability of a Pd/C catalyst was measured by taking 2.2 nm Pd NPs for an example. Prior to each cycle of FAD measurement, the catalyst was rinsed with a large amount of ultrapure water, then dried at 343 K and dispersed in a fresh FA-SF solution at 303 K. Figure 6 show the volume of hydrogen production with elasping time for three consecutive runs under otherwise same conditions. The ICP-AES analysis of the catalyst indicates the loading of Pd is 4.9 wt.% and 4.8 wt.% after the 1st and 2nd runs, respectively, nearly identical to the initial one given the uncertainty in the ICP-AES measurement. It turns out that FAD activity on this catalyst remain reasonably stable with increasing cycle. The average mass and specific activities for a given Pd/C sample in a given FA-SF solution at 303 K are largely dependent on the length of elasping time used for the evaluation. Herein, those evaluated based on 2 min, 5 min and 8 min after FAD were respectively used to represent approximately hydrogen production rates per unit of Pd mass (defined as the mass activity) and TOF values covering initial, half-course and near-finish stages of FAD, as shown in Table 2 and Figure 5c-d. The mass activity is found to increase monotonically with decreasing size of Pd NPs no matter which FAD stage is concerned, promising for even smaller sized Pd NPs in practice. Nevertheless, on the basis of cuboctahedron model, the three TOF vs. size plots suggest that the TOF value increases as the size of Pd NPs decreases from 5.2 to 2.8 nm with a steep rise of TOF from 3.5 down to 2.8 nm, and the increase rate of TOF slows down with further decreasing the Pd particle size, or even levels off for the TOF evaluated over 8 min of FAD.

11



1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 12 of 29

Figure 7. Anodic COads stripping voltammograms for Pd/C catalysts recorded in 0.5 M HClO4 solution at a scan rate of 10 mV s-1. (a) The oxidative CO removal region of different Pd sizes. (b) The full voltammetric curve shown for the 2.2-nm Pd/C catalyst. In literature, CO adsorption in gas atmosphere was often used to estimate exposed Pd surface sites. 26, 75 However, this raises the concern on possible measurement error since the actual FAD occurs in a solution rather in a gas phase. Alternatively, TOF values can be determined on the basis of electrochemical surface area (ECSA), that is,

TOF=

$%&' %( )*+(,-' ,-./0' )/.' 1

2345 ∙ $7 ∙ 8 9:

where NA is the Avogadro constant and ρ is the surface packing density of Pd atoms, 1×1019 m-2 . 39 The ECSA of a Pd/C catalyst was calculated by anodic stripping of a predosed CO monolayer on Pd NPs (see Figure 7(a)) using the following equation: ;

Surfactant-Free Synthesis of Carbon-Supported Palladium

Recommend Documents