Bioethanol to hydrogen membrane surface characteristics change study

Бюллетень науки и практики / Bulletin of Science and Practice

UDC 544.47                                         

Green and sustainable economy determine essential demand for green energy production [1, 2]. One possible way to solve this problem is to develop method for hydrogen production from renewable resources [3, 4]. Green hydrogen can be produced using water electrolysis and bio renewable resources catalytic pyrolysis, gasification and transformation [5]. Bioethanol can be considered as bio renewable resource easily produce from agriculture and wood processing wastes and applicable for hydrogen production [6, 7]. Bioethanol can be converted into hydrogen using ethanol water gas shift reaction. Efficiency of the process is determined catalysts activity and selectivity [8]. Typically, Pt, Pd, Ni, Fe, Cu, Zn, Al containing catalysts are used to provide this process [9]. Catalytic process of bioethanol to hydrogen transformation can be combined with membrane proses of hydrogen separation [10]. The main problem of such systems is catalyst deactivation due to surface carbonization and active metals degradation [11]. Therefor insight in surface change during catalytic membrane operation is essential for understanding ways to decrease membranes deactivation.

Nitrogen physisorption can be considered as effective method for obtaining membranes surface characteristics and determination of surface change during membrane work. Typically, nitrogen physisorption measurements are provided using static volumetric method, were nitrogen adsorption determined by mass balance equations and pressure drop in sample sell. For evaluation of obtained data several models can be used including Langmuir, BET, BJH [12] and t-plot methods can be used. Langmuir model includes adsorption of one molecular layer on solid surface (1).

p _ 1 p

V^+й™

Where p — system pressure, Pa; V a — volume of adsorbed gas; V m — volume of adsorbed gas forms monomolecular layer; b — correlation coefficient.

Most frequently used Brunauer, Emmett, and Teller model includes possibility of gas multilayer condensation (2).

_Л_ = ±+£-1Й(2)

v_a(p_o-p)  v_mc  V_mc\pj

Where P 0 — saturation pressure of gas; C — gas adsorption constant;

Barrett-Joyner-Halenda (BJH) method is based on Kalvin equation (3).

^In^U'C

VP

Where ν — molar volume; σ — liquid surface tension; C — is mean curvature defined on formula 4.

C =- +

Г 1

^ 2

Where r — interfacial curvature;

The equation 4 can be converted into equation 5.

p    2yvcos6

ln^=-^^

Where ɣ — interfacial tension, θ — contact angle of condensed liquid with wall.

t-plot model uses equation (6) to evaluate micropores volume, external surface area and micropore surface area.

t=

J

13.99 log 1o (y)+0.034

wheret — film thickness;

Application of discussed models for surface characterization gives possibility for evaluation of membranes surface degradation during its work.

Materials and Methods

Specific surface area and porosity were determined using the following instruments: Becman coulter SA3100 surface area and pore size distribution analyzer (Coulter corporation, Miami, Florida), sample preparation device: Becman coulter SA-prep (Coulter corporation, Miami, Florida). For analysis, the sample is placed in a pre-weighed quartz cuvette, which is installed in the SA-PREP™ sample preparation instrument. Sample preparation parameters: temperature — 120℃; gas — nitrogen; preparation time — 60 min. After completion of sample preparation, the cuvette is cooled and weighed, and then transferred to the analytical port of the BECMAN COULTER SA 3100™ instrument. Gas was sequentially supplied to the analytical cell and the equilibrium pressure in the system was determined.

Results and Discussions

Initial membrane characterized by Langmuir surface area 1.3 m2/g, BET surface area 2.9 m2/g, t-plot surface area 25.3 m2/g. BJH-pore volume distribution (Figure 1) shows that most part of mesopores have diameter less than 6 nm.

0,005

0,0045   •

0,004

0,0035

0,003

0,0025

0,002

0,0015

0,001

0,0005 0

-10       10       30       50       70       90       110      130

Avarage diameter, nm

Figure 1. Pore volume distribution for initial inorganic membrane

Membrane operation for 50 h on stream results in increase of Langmuir surface area 4.7 m2/g, BET surface area up to 12.3 m2/g, t-plot surface area 78.4 m2/g. BJH-pore volume distribution (Figure 2) shows shift of average pore diameter to 20 nm.

0,0045

0,004

0,0035

0,003

0,0025

0,002

0,0015

0,001

0,0005

0        20       40        60        80       100

Avarage diameter, nm

Figure 2. Pore volume distribution for membrane from ethanol to hydrogen reaction operation for 50 h on stream

Membrane operation for 100 h on stream results in increase of Langmuir surface area 7.4 m2/g, BET surface area up to 24.9 m2/g, t-plot surface area 53.9 m2/g. BJH-pore volume distribution (Figure 3) shows small shift of average pore diameter to region less than 15 nm.

0,004

0,0035

0,003

0,0025

0,002

0,0015

0,001

0,0005

0        20        40

♦■

Avarage diameter, nm

Figure 3. Pore volume distribution for membrane from ethanol to hydrogen reaction operation for 100 h on stream

Membrane surface degradation during methanol to hydrogen reaction process shows partial increase of membrane mesoporosity and microporosity that can be explained by partial hydrolysis of membrane surface with water steam forming during ethanol to hydrogen transformation process.

Conclusions

Membrane surface degradation is a complex problem for inorganic catalytic membranes. Ethanol to hydrogen catalytic transformation process provided on inorganic membranes strongly suffers from membrane degradation because of high exothermic effect and high amount of water forming during the process. After 50 hours on stream Langmuir surface area increase from 1.3 m2/g to 4.7 m2/g, BET surface area increase form 2.9 m2/g to 12.3 m2/g and t-plot surface area increase from 25.3 m2/g to 78.4m2/g. After 100 hours on stream Langmuir surface area increase to 7.4 m2/g, BET surface area increase to 24.9 m2/g and t-plot surface area decrease to 53.9m2/g.

The reported study was funded by Russian Foundation for Basic Research (RFBR), project number 20-08-00433 А.