Electroless Pd membrane deposition on alumina modified porous Hastelloy substrate with EDTA-free bath

Electroless Pd membrane deposition on alumina modified porous Hastelloy substrate with EDTA-free bath

international journal of hydrogen energy 35 (2010) 2328–2335 Available at www.sciencedirect.com journal homepage: www.elsevier.com/locate/he Electr...

1MB Sizes 1 Downloads 7 Views

Recommend Documents

No documents
international journal of hydrogen energy 35 (2010) 2328–2335

Available at www.sciencedirect.com

journal homepage: www.elsevier.com/locate/he

Electroless Pd membrane deposition on alumina modified porous Hastelloy substrate with EDTA-free bath Shin-Kun Ryi a,b,*, Nong Xu a, Anwu Li a,b,**, C. Jim Lim a,b, John R. Grace a,b a b

Department of Chemical and Biological Engineering, University of British Columbia, 2360 East Mall, Vancouver, BC, Canada V6T 1Z3 Membrane Reactor Technologies Ltd. 6301 Stadium Road Vancouver, BC, Canada V6T 1Z4

article info


Article history:

This study demonstrates palladium membranes can be electrolessly plated on aluminum

Received 30 November 2009

oxide-modified porous Hastelloy with hydrazine using an EDTA-free bath. The plating bath

Received in revised form

temperature affected the membrane surface morphology, with the palladium grain size

10 January 2010

increasing with increasing temperature. A 7.5 mm thick membrane plating was obtained at

Accepted 15 January 2010

room temperature. Helium leak testing confirmed that the membrane was free of defects.

Available online 4 February 2010

Hydrogen permeation test showed that the membrane had a hydrogen permeation flux of


There was no measurable interdiffusion between the membrane film and the porous


Hastalloy substrate at 823 K. This room temperature membrane plating method provides


several advantages such as very high selectivity, stability, favorable energy efficiency and



3.3  101 mol m2 s1 at a temperature of 823 K and at a pressure difference of 100 kPa.

Electroless plating

ª 2010 Professor T. Nejat Veziroglu. Published by Elsevier Ltd. All rights reserved.

Porous substrate



The demand for ultra-high-purity hydrogen has increased rapidly recently for several industrial processes such as chemical synthesis, hydrogenation, semiconductor manufactures, and fuel cells. Hydrogen can be produced through a number of processes [1]. Commercially available hydrogen typically contains impurities such as CO, CO2, O2, N2, H2O, and CH4, which must be separated from the hydrogen. The major challenges for more widespread use of hydrogen and a hydrogen-based economy are the development of low cost technologies for its production, separation, and purification. In the face of these challenges, hydrogen purification (separation) is a dynamic and rapidly growing field. Hydrogen can be purified through several techniques, such as pressure swing adsorption (PSA), cryogenic distillation, getters,

and membrane separation. Membrane-related processes are one of the most promising technologies for producing of highpurity hydrogen [2–5]. Among the several kinds of membranes, Pd and Pd-based dense membranes command great attention due to their very high hydrogen selectivity, high hydrogen permeability, and chemical compatibility with hydrocarboncontaining gas streams. In view of these benefits, Pd-based membranes have great commercial potential compared with other hydrogen-selective membranes [6–9]. Composite structures consisting of ultra-thin membrane film and porous substrate can successfully reduce a material costs and provide high hydrogen permeation rates. The materials which have been applied commercially for porous substrates include ceramics, glass and metals such as stainless steel (PSS), Inconel and nickel [1]. For practical applications, porous metals have the advantages of facilitating scale-

* Corresponding author. Tel.: þ1 604 827 5563; fax: þ1 604 822 6003. ** Corresponding author. Tel.: þ1 604 827 4343. E-mail addresses: [email protected], [email protected] (S.-K. Ryi), [email protected], [email protected] (A. Li). 0360-3199/$ – see front matter ª 2010 Professor T. Nejat Veziroglu. Published by Elsevier Ltd. All rights reserved. doi:10.1016/j.ijhydene.2010.01.054

[12] [13] [14] [15] [16] [17] [18] [19] [20] This work 18.8 14.4 ND 13.3 ND ND 7.0 ND 20.6 20.1 0.5 0.6 0.5 1 1 0.5 1 0.5 0.5 0.5 4.1  109 1.1  108 1.7  108 2.6  109 5.7  109 6.5  109 1.1  108 2.1  108 1.6  108 1.6  108 1.9  108 333 r.t. (w298) r.t. 323 333 r.t. (296) 313 323 333 r.t. (w293)

8 10.5 19 ND 1 >6 ND ND ND 15

773 753 773 723 773 780 773 723 823 773 823

160 100 100 100 140 100 100 50 100 100 100

1.0  101 3.0  101 2.5  101 2.3  101 2.5  101 1.7  101 2.8  101 0.79  101 4.3  101 2.8  101 3.3  101

>1  107 (H2/N2) N (H2/N2) 800 (H2/N2) 3115 (H2/N2) N (H2/N2) 600 (H2/N2) >3000 (H2/N2) 608 (H2/N2) N (H2/N2) N (H2/He)

Ea [kJ mol1] n-Value Selectivity H2 permeability [mol m1 s1 Pa0.5] Plating time [h]

Temp. [K]

DP [kPa]

H2 flux [mol m2 s1]

8 5 9 1.5 2–4 5 5 19 5 7.5

Palladium membranes were synthesized on pre-treated porous Hastelloy discs of 25.4 mm diameter and 1.2 mm thickness (Mott, 0.2 mm grade). The pre-treatment procedure followed the method described by Li et al. [26], including polishing and etching, followed by diffusion barrier coating with aluminum oxide sol particles of two different mean sizes, i.e. 2.5 and 0.3 mm. The palladium layer deposition process included three steps: Pd nuclei seeding, electroless plating and post-plating

Pd/PSS PdCu/ZrO2/PSS Pd/TiO2/PSS Pd/Al2O3 Pd/PSS Pd/Al2O3 Pd/Al2O3 Pd/NaAZ/PSS Pd/Al2O3/PSS Pd/Al2O3/PHA

Membrane preparation

Plating temp. [K]


Thick. [mm]




Table 1 – Comparison of hydrogen permeation performance for Pd-based composite membrane prepared by electroless plating.

up to commercial unit. However, atomic interdiffusion of metals between the thin Pd/Pd alloy layer and the metal components occurs during high-temperature processing. To inhibit atomic interdiffusion, a ceramic layer must be introduced as a diffusion barrier. Common practices to form an interdiffusion barrier layer are coating a thin ceramic layer such as ZrO2, TiO2, Al2O3 (see Table 1) and SiO2 [10]. Common methods for preparing Pd-based composite membranes are sputtering, chemical vapor deposition (CVD), electroplating, electroless plating and spray pyrolysis [1]. Auto catalytic deposition, so-called electroless plating, has a number of advantages over other methods, such as uniformity of deposits on complex shapes, hardness of the deposits, low cost and simplicity [11]. Membranes made by electroless plating have the highest permeabilities compared with other methods such as CVD and sputtering [12]. Most electroless palladium plating studies have used EDTA salt because EDTA increases the stability of the bath, assisting in controlling the plating process at different temperatures [13–21]. However, EDTA results in low purity of the palladium layer owing to incorporation of the EDTA complex within the metal deposit [22], moreover, EDTA can be a source of carbon which degrades membrane performance [12,23–25]. Therefore, the group of Way has developed an electroless palladium plating method which avoids EDTA when making thin unsupported palladium alloy foil membranes [24,25]. They did not show how to increase the stability of the plating bath. They added excess ammonium hydroxide into hydrazine. Even though the deposition rate is greater in a higher temperature bath, the plating bath is not stable for sufficient time, causing bulk precipitation. This precipitation starts quickly at higher temperatures, requiring a fresh plating bath solution at frequent intervals [18,22]. The bulk precipitation can also reduce the palladium deposition yield on substrates. This study was undertaken to optimize electroless palladium plating with an EDTA-free bath to avoid carbon deposition and palladium bulk precipitation. Porous Hastelloy (PHA) was used as the support. This porous Hastelloy was pretreated by means of the method developed by Membrane Reactor Technologies (MRT Ltd.) involving polishing and etching, followed by coating with two aluminum oxide layers [21,24]. The plating bath temperature was varied from 293 to 308 K to determine the effect of temperature on the membrane morphology. Hydrogen permeation and helium leak tests were employed to test the performance of the resulting palladium layers.


international journal of hydrogen energy 35 (2010) 2328–2335



international journal of hydrogen energy 35 (2010) 2328–2335

treatment. Prior to electroless plating, Pd nuclei were seeded by impregnating the substrate with a 0.1 M PdCl2 solution, followed by drying and reduction under hydrogen at 723 K. Electroless Pd plating was then carried out in an EDTA-free plating bath whose composition is given in Table 2. The influence of plating temperature on membrane morphology was tested with small pieces of alumina-coated Hastelloy at temperature between 293 and 308 K, with 293 K chosen for the membrane fabrication. Plating was repeated five times, each with 6 ml plating solution and 3 h plating time. 6 ml plating solution was large enough for discs of 25.4 mm diameter and 1.2 mm thickness. 15 h were required to achieve a 7.5 mm thickness palladium layered membrane. It is very similar with other studies [14,15] using EDTA at room temperature. After completing the plating, the membrane was cleaned several times with deionized water and heat-treated. During the heat-treatment, helium was introduced at a temperature of 573 K, and then hydrogen replaced the helium. The heat-treatment was continued for 5 h at 723 K. This heat-treatment is not for elimination of defects, but for crystallization of palladium grain to enhance hydrogen permeation. Tucho et al. [27] observed enhancement of hydrogen permeation for thin Pd–Ag membranes after heat-treatment with inert gas. They concluded that the permeation enhancement may result from beneficial conditions for atomic diffusion leading to other microstructural changes in addition to surface segregation. The membranes were characterized by scanning electron microscopy (SEM), energy dispersive X-ray spectroscopy (EDX) and thermogravimetric analysis (TGA).


Table 2 – Composition of electroless plating bath. Components PdCl2 NH4OH (28%) HCl N2H4 (1%) pH

Concentration or value 3.2 g/L 320 ml/L 4.0 ml/L 200 ml/L w11

Permeation measurements

Prepared membranes were mounted in stainless steel permeation cells with graphite gaskets. Permeation tests were then conducted using hydrogen at temperatures in the range 673– 823 K, with a transmembrane pressure difference of 40– 100 kPa. Helium leak tests were conducted at a transmembrane pressure difference of 100 kPa and room temperature before permeation tests and 823 K after permeation tests. The membranes were heated slowly under helium until the temperature exceeded 573 K to prevent phase transition causing membrane fracture [28]. The permeation rate of hydrogen was measured by a soapbubble flow meter. When testing was complete, membranes were slowly cooled under helium, followed by characterization of the surface morphology, elemental composition and membrane thickness using SEM/EDX.


Results and discussion


Electroless plating

To determine the effect of temperature on the plating behavior, palladium was plated on small pieces of alumina modified substrate at three temperatures, i.e. 293, 303 and 308 K. Previously reported results [18] suggest that lower temperatures reduce grain growth and non-uniformities during electroless plating, leading to better quality membranes which are more compact and have fewer defects. Fig. 1 shows surface SEM images after plating at all three temperatures. It can be seen that

Fig. 1 – SEM images after plating at different temperatures: (a) 293 K; (b) 303 K; (c) 308 K.


international journal of hydrogen energy 35 (2010) 2328–2335

the bulk grains of 308 K bath were smaller than those of other baths but the surface grain size was reduced and more uniform with decreasing plating temperature, as in the previous study [18]. McCool and Lin showed that a Pd–Ag membrane of submicron thickness and larger grain size has greater hydrogen permeability than smaller-grained Pd–Ag membranes [29]. In their study, membranes were sputtered on alumina supports and heat-treated to increase grain sizes. They concluded that narrower grain boundaries resulted from grain expansion at higher temperature, leading to improved hydrogen permeability. Souleimanova et al. [13] deposited Pd membranes on Vycor glass and porous stainless steel substrates by osmosis electroless plating with polyethylene glycol as the osmotic solution. Palladium grains from the osmosis method were finer than from the conventional method. They concluded that a finer grain microstructure leads to higher hydrogen permeability and superior thermal stability. Room temperature was adopted electroless plating after finding that a higher temperature (308 K) plating bath was unstable enough that the plating solution became murky within 30 min indicating bulk precipitation. Surface and cross-sectional SEM images of a palladium composite membrane prepared in a 293 K bath are reproduced in Fig. 2. The cross-sectional SEM image was taken after

permeation test. The surface morphology of the membrane is similar to that in Fig. 1(a), with a uniform palladium grain size of 1–2 mm. This indicates that the plating method was reproducible. It can be seen that the surface of this membrane was very smooth, and there were no visible defects on the membrane surface. The cross-sectional SEM image in Fig. 2(b) indicates that the palladium layer thickness was ca. 7.5 mm. The rate and yield of palladium deposition on the pre-treated substrate were determined for a 6 ml plating solution at room temperature. Fig. 3 shows that palladium deposition increased as a function of the plating time and around 80% palladium was utilized after 3 h. The deposition rate slightly decreased with plating time as in previous studies [18, 22], probable due to concentration and depletion of a decrease of Pd(NH3)2þ 4 hydrazine in the plating solution [18]. Catalytic decomposition of hydrazine by palladium may be a factor [30]. Murkiness of the bath is an indicator [18]. In our case the plating bath did not become murky at room temperature. Volbe et al. [22] showed that their Pd deposition rate increased to w3.0  104 g cm2 after 30 min at 333 K for a Pd plating bath containing EDTA as a stabilizer. Nair et al. [18] found that the palladium deposition rate depends on temperature and on the N2H4 concentration. Higher bath temperature and higher Pd and hydrazine concentrations resulted in more rapid growth and earlier onset of bath decomposition, i.e. bulk deposition. They achieved a maximum deposition rate of 1.4 mm h1 after several minutes and an average rate of 0.75 mm h1 at




Pd weight, g cm


1.4e-3 1.2e-3 1.0e-3 8.0e-4 6.0e-4 4.0e-4 2.0e-4 0.0 0.0








Time, h


b Plating yield, %





0 0.0

Fig. 2 – Surface and cross-sectional SEM images of membrane prepared by electroless plating using EDTA-free bath at 293 K.








Time, h

Fig. 3 – Palladium deposition rate (a) and plating yield (b) as function of plating time at 293 K.

international journal of hydrogen energy 35 (2010) 2328–2335

a temperature of 296 K with 28 mM Pd and 110 mM hydrazine and no bulk deposition for 5 h. In our case, we obtained a palladium weight gain of 5.0  104 g cm2 after 30 min and a 3 h-average deposition rate of 0.4 mm h1 at 293 K for a solution containing 18 mM Pd and 62.5 mM hydrazine without bulk deposition. Volpe et al. [22] showed the presence of Na associated with incorporation of EDTA by using EDX and carboxyl (band at 1655 cm1) and amino (bands at 1387 and 1319 cm1) groups from EDTA based on FTIR. Roa and Way [31] reported a significant proportion of carbon (w6.5 wt%) from the EDTA agent in electroless-plated Pd and Pd–Cu membranes by TGA analysis under flowing CO2. Carbon removal only started when the sample reached about 160  C and proceeded rapidly from there until it reached 400  C. Fig. 4 shows the EDX (a) of fresh membrane and TGA (b) analysis under flowing CO2. A sample for TGA analysis was made by peeling Pd foil from the substrate, and additional plating was carried out to gain enough Pd for TGA analysis. The sample (ca. 9.89 mg) was ramped to 400  C at 1  C min1, held at 400  C for 5 h, and then cooled at 1  C min1 to room temperature. The EDX profile of fresh membrane indicates that there no other elements except Pd were present in our membrane. The TGA analysis shows that the weight difference was less then 0.3 wt% indicating there no carbon contaminant was present in our membranes.


Hydrogen permeation

Gas permeation tests were conducted with pure hydrogen at temperatures of 673–823 K and at pressure differences across the membrane of 40–100 kPa. The hydrogen fluxes for the composite membranes are plotted in Fig. 5. As expected the hydrogen flux increased with increasing temperature and feed side pressure. The hydrogen flux increased to 3.3  101 mol m2 s1 at a temperature of 823 K and a pressure difference of 100 kPa. Table 1 compares the permeation performance of our membrane with other reported composite membranes prepared by electroless plating on various porous substrates. It can be seen that Pd/ceramic and Pd/ceramic/

0.35 673 K 723 K 773 K 823 K

0.30 H2 permeation flux, mol m-2 s-1


0.25 0.20 0.15 0.10 0.05 0.00 0







Pressure difference, kPa

Fig. 5 – Hydrogen permeation flux as a function of pressure difference across the membrane.

metal membranes have higher hydrogen permeability than most other Pd/metal composite membranes. Our membrane has a similar hydrogen permeation flux and superior selectivity relative to other electrolessly plated membranes. The advantages of our membrane fabrication procedures are its excellent selectivity, energy efficiency and very simple equipment. Our membranes can be fabricated at room temperature and, except for the plating bath, do not need other equipment like a water bath required by other hightemperature electroless plating methods. Hydrogen diffuses through palladium-based alloy membrane via a solution and diffusion mechanism. This phenomenon is frequently described by: JF ¼

 QF  n Pup  Pndown l


where JF and QF denote the hydrogen flux (mol m2 s1) and permeability (mol m m2 s1 Pan) for palladium alloy membrane (film), respectively, and l is the thickness (m) of the

Fig. 4 – EDX (a) and TGA (b) profiles of membrane prepared by electroless plating using EDTA-free bath at 293 K: The sample was ca. 9.89 mg for TGA analysis; The TGA analysis was carried out under flowing CO2.


international journal of hydrogen energy 35 (2010) 2328–2335

  Ea Q ¼ Q0 exp RT


or   Q Q0 Ea exp ¼ l RT l

7 Original PHA Alumina modified PHA

6 H2 permeation flux, mol m-2 s-1

palladium-based alloy layer. Pup and Pdown are the pressures on the feed and permeate sides, respectively. If the diffusion of hydrogen atoms through the dense metal layer is ratelimiting, the hydrogen flux, then n ¼ 0.5, as in the well-known Sieverts’ law. In our work, QF and n-value were calculated from the overall hydrogen flux, rather than the hydrogen flux through the palladium layer, because the membranes were composite, consisting of a dense film and a porous support, separated by alumina. Fig. 6 shows the dependency of the hydrogen flux on the feed side pressure for the prepared composite membrane at various temperatures. As shown in Fig. 6 the pressure exponent of 0.5 gave good fits. Hydrogen transport through a dense Pd-based alloy membrane is an activated process. The relationship between hydrogen permeability and temperature can be described by an Arrhenius law:

5 4 3 2 1 0 0






Pressure difference, MPa

Fig. 7 – Hydrogen permeation flux of original porous Hastelloy and pre-treated support as a function of pressure difference across them.


where Q is hydrogen permeability (mol m m2 s1 Pan), l is the membrane thickness (m), Ea the activation energy (kJ mol1), R the universal gas constant (8.314 J mol1 K1) and T the temperature in Kelvin. Activation energies for Pd-based membranes have been reported from 5.4 to 38 kJ mol1 [1,32]. There are few data showing the effects of different factors on the activation energy. The fabrication method appears to influence permeation in ultra-thin Pd-based membranes, probably due to differences in microstructure or impurities, on the surface, in the bulk, or at grain boundaries [33]. In our membrane, the hydrogen permeation flux of the pre-treated porous Hastelloy was 50% of that for the original one as shown in Fig. 7. Hydrogen permeance of the pre-treated substrate was 7.5  105 mol m2 s1 Pa1 larger than the value from our previous modified stainless steel [26] of w4.8  105 mol m2 s1 Pa1. This value is more than 22

times higher than for the Pd composite membrane at 823 K indicating that the mass transfer resistance of the support is negligible compared to that of Pd layer. Nam and Lee [10] attempted to prepare a thin Pd-based composite membrane on modified porous stainless steel (0.5 mm grade) with submicron nickel and silica. The hydrogen permeance of their pre-treated substrate was only 1.0  105 mol m2 s1 Pa1, w13% of that of our pre-treated substrate suggesting that our pre-treatment method is superior. The activation energy of the membranes evaluated from an Arrhenius plot of overall hydrogen permeability against the reciprocal temperature in Fig. 8 was from 20.1 to 22.1 kJ mol1 for pressure differences of 100–40 kPa, similar to values for other Pd-based membranes [13,21,34]. In principle, activation energy is independent of H2 pressure. The variation of activation energy might be experimental deviation.

673 723 773 823




ln (H2 permeance, mol m-2 s-1 Pa-0.5)

H permeation flux, mol m-2 s-1 2




100 kPa 80 kPa 60 kPa 40 kPa





0.0 0







Pup0.5 - Pdown0.5, Pa0.5

Fig. 6 – Hydrogen permeation flux as a function of difference between square roots of pressure at various temperatures.


-6.8 1.20







103/T, 103 K-1

Fig. 8 – Arrhenius plot of overall hydrogen permeability as a function of reciprocal temperature.


international journal of hydrogen energy 35 (2010) 2328–2335

that no defects were present. After testing, the membrane was removed out from the testing module and maintained in hydrogen at 823 K for 100 h. Fig. 10 shows an EDX line-scan after the heattreatment. After stability test at 823 K, the membrane was cooled in helium so that no metal-hydride could be formed. Interdiffusion between the membrane film and porous Hastelloy substrate was negligible because the pre-treated aluminum oxide layer was effective in preventing interdiffusion.


H2 permeation flux, mol m-2 s-1

773 K

823 K






0.0 0






Time, h

Fig. 9 – Hydrogen permeation flux as a function of time of operation at a pressure difference of 100 kPa.


Membrane stability

Hydrogen permeation fluxes were measured continuously for 24 h at temperatures of 773–823 K and at a pressure difference of 100 kPa. As shown in Fig. 9, the hydrogen permeation flux remained constant over this period. After the permeation test, helium leak testing was carried out at 823 K again at a pressure difference of 100 kPa. This helium leak testing confirmed


Hydrogen permeation tests and membrane stability tests for palladium composite membranes deposited on aluminum oxide-modified porous Hastelloy with hydrazine method using an EDTA-free bath revealed that:  Surface morphology was affected by the plating bath temperature. Palladium grain size decreased with decreasing plating temperature.  Membranes prepared at 293 K were very compact with a dense film layer an a uniform grain size of w1–2 mm.  Membranes were confirmed to be defect-free by SEM analysis and helium leak testing. The hydrogen permeation flux was 3.3  101 mol m2 s1 at 823 K for a pressure difference across the membrane of 100 kPa.  Membranes were stable, without any measurable changes in morphology or interdiffusion between membrane film and the porous Hastelloy substrate at a temperature of 823 K, indicating that the modified aluminum oxide layer acted as an effective diffusion barrier.  The new method provides high selective membrane, is energy-efficient and requires only very simple equipment relative to other reported electroless palladium plating methods.

Acknowledgements This work was supported by Natural Science and Engineering Research Council of Canada, Membrane Reactor Technologies and Noram Engineering and Constructors. Dr. Shin-Kun Ryi was supported by the Korea Research Foundation Grant funded by the Korean Government [KRF-2008-357-D00054].


Fig. 10 – Sampling line (a) for EDX line-scanning profile (b) of membrane surface after stability test.

[1] Ryi S-K. The study of Pd–Cu–Ni ternary alloyed hydrogen membranes deposited on porous nickel supports, Doctoral thesis, Korea University 2007. [2] Phair JW, Badwal SPS. Materials for separation membranes in hydrogen and oxygen production and future power generation. Sci Technol Adv Mater 2006;7:92–805. [3] Report of the DOE Workshop on hydrogen separations and purification, Arlington, VA, September 8–9, 2004. [4] Uemiya S. Brief review of steam reforming using a metal membrane reactor. Topics Catal 2004;9:9–84.

international journal of hydrogen energy 35 (2010) 2328–2335

[5] Dolan MD, Dave NC, Ilyushechkin AY, Morpeth LD, McLennan KG. Composition and operation of hydrogenselective amorphous alloy membranes. J Membr Sci 2006;285: 30–55. [6] Pandey P, Chauhan RS. Membrane for gas separation. Prog Polym Sci 2001;26:853–93. [7] Ryi S-K, Park J-S, Kim S-H, Kim D-W, Woo B-I, Grace JR. Development of a hydrogen purifier with Pd-based composite membrane. Korean J Chem Eng 2010;27:235–40. [8] Armor JN. Applications of catalytic inorganic membrane reactor to refinery products. J Membr Sci 1998;147:217–33. [9] Adhikari S, Fernando S. Hydrogen membrane separation techniques. Ind Eng Chem Res 2006;45:875–81. [10] Nam S-E, Lee K-H. Hydrogen separation by Pd alloy composite membrane: introduction of diffusion barrier. J Membr Sci 2001;192:177–85. [11] Li A, Liang W, Hughes R. Fabrication of dense palladium composite membranes for hydrogen separation. Catal Today 2000;56:45–51. [12] Lu GQ, Diniz da Costa JC, Duke M, Giessler S, Socolow R, Williams RH, et al. Inorganic membranes for hydrogen production and purification: a critical review and perspective. J Colloid Interface Sci 2007;314:589–603. [13] Souleimanova RS, Mukasyan AS, Varma A. Pd membranes formed by electroless plating with osmosis: H2 permeation studies. AIChE J 2002;48:262–8. [14] Gao H, Lin JYS, Li Y, Zhang B. Electroless plating synthesis, characterization and permeation properties of Pd–Cu membranes supported on ZrO2 modified porous stainless steel. J Membr Sci 2005;265:142–52. [15] Huang Y, Dittmeyer R. Permeation and characterization of composite palladium membranes on sinter-metal supports with a ceramic barrier against intermetallic diffusion. J Membr Sci 2006;282:296–310. [16] Sun GB, Hidajat K, Kawi S. Ultra thin Pd membrane on a-Al2O3 hollow fiber by electroless plating: high permeance and selectivity. J Membr Sci 2006;284:110–9. [17] Shi Z, Wu S, Szpunar JA, Roshd M. An observation of palladium membrane formation on a porous stainless steel substrate by electroless deposition. J Membr Sci 2006;280:705–11. [18] Nair BKR, Choi J, Harold MP. Electroless plating and permeation features of Pd and Pd/Ag hollow fiber composite membranes. J Membr Sci 2007;288:67–84. [19] Zhang X, Xiong G, Yang W. A modified electroless plating technique for thin dense palladium composite membranes with enhanced stability. J Membr Sci 2008; 314:226–37. [20] Bosko ML, Ojeda F, Lombardo EA, Cornaglia LM. NaA zeolite as an effective diffusion barrier in composite Pd/PSS membranes. J Membr Sci 2009;331:57–65.


[21] Li A, Grace JR, Lim CJ. Preparation of thin Pd-based composite membrane on planar metallic substrate, Part II. Preparation of membranes by electroless plating and characterization. J Membr Sci 2007;306:159–65. [22] Volpe M, Inguanta R, Piazza S, Cunseri C. Optimized bath for electroless deposition of palladium on amorphous alumina membranes. Surf Coat Technol 2006; 200:5800–6. [23] Collins JP, Schwartz RW, Sehgal R, Ward TL, Brinker CJ, Hagn GP, et al. Catalytic dehydrogenation of propane in hydrogen permselective membrane reactors. Ind Eng Chem Res 1996;35:4398–405. [24] Thoen PM, Roa F, Way JD. High flux palladium–copper composite membranes for hydrogen separations. Desalination 2006;193:224–9. [25] Gade SK, Thoen PM, Way JD. Unsupported palladium alloy foil membranes fabricated by electroless plating. J Membr Sci 2008;316:112–8. [26] Li A, Grace JR, Lim CJ. Preparation of thin Pd-based composite membrane on planar metallic substrate, Part I: pretreatment of porous stainless steel substrate. J Membr Sci 2007;298:175–81. [27] Tucho WM, Venvik HJ, Stange M, Walmsley JC, Holmestad R, Bredesen R. Effects of thermal activation on hydrogen permeation properties of thin, self-supported Pd/Ag membranes. Sep Purif Technol 2009;68:403–10. [28] Katsnel’son AA, Knyazeva MA, Revkevich GP. Characteristics of the evolution of defect structure during phase transformations in the Pd–H system. Phys Solid State 1997; 38:1132–8. [29] McCool BA, Lin YS. Nanostructured thin palladium–silver membranes: effects of grain size on gas permeation properties. J Mater Sci 2001;36:3221–7. [30] Mallory GO, Hajdu JB. Electroless plating: fundamentals and applications. In: Okinaka Y, Wolowodiuk C, editors. Electroless plating of platinum group metals. New York: William Andrew Publishing; 1990. p. 421–40. [31] Roa F, Way JD. The effect of air exposure on palladium–copper composite membranes. Appl Surface Sci 2005;240:85–104. [32] Ryi S-K, Park J-S, Kim S-H, Kim D-W, Cho K-I. Fromation of a defect-free Pd–Cu–Ni ternary alloy membrane on a polished porous nickel support (PNS). J Membr Sci 2008;318: 346–54. [33] Ward TL, Dao T. Model of hydrogen permeation behavior in palladium membranes. J Membr Sci 1999;153:211–31. [34] Tong J, Su L, Haraya K, Suda H. Thin Pd membrane on a-Al2O3 hollow fiber substrate without any interlayer by electroless plating combined with embedding Pd catalyst in polymer temperate. J Membr Sci 2008;310:93–101.