Hydrothermal stability of a new hybrid membrane in dehydration applications

Transcription

1 Hydrothermal stability of a new hybrid membrane in dehydration applications Jaap Vente

2 Approach Four way competences: Materials research System development Process design Implementation facilitation Five different lines of applications Pervaporation NH 3 separation Oxygen production Hydrocarbons separations Hydrogen purification

3 Why dehydration with membranes? By 2015 potential energy savings: NL: 7 PJ/yr (2% of industrial energy consumption) World: 240 PJ/yr Production of e.g. Esters and resins Methyl ethyl ketone (MEK) Ethanol Ureum Acrylic acid

4 Industrial performance demands Standard ECN tests: 5 wt.% water in butanol Fluxes 5 kg/m 2 h Selectivity > 220 Longevity 2-3 years Conditions: Various ph Temperatures up to 150 C

5 Tubular microporous membranes 1000 nm Pores < 1nm ZrO2 /TiO 2 4 nm pores 120 nm pores

6 Materials covered Previous developments: SiO 2 Methylated SiO 2 New leads: TiO 2 ZrO 2 Hybrid silica, organic bridges (Ceramic supported polymers)

7 SiO 2 Me-SiO 2 pervaporation results (95ºC) J H2O /kg h -1 m SiO 2 1 step Me-SiO 2 10% MTES 30% MTES t/d 300 t/d wt.% H 2 O in BuOH, 10 mbar Addition of MTES gives better performance with time Constant performance possible for over 18 months! Published data Chem.Comm.2004,

8 SiO 2 Me-SiO 2 pervaporation up to 165ºC J H2O / kg h -1 m -2 ) C 135 C 165 C single step Me-SiO wt.% H 2 O in BuOH, 10 mbar Failure within weeks No clear relation with temperature t/d J BuOH / kg h -1 m -2 ) t / d 30 40

9 Membrane surface after testing (Me-SiO 2 ) Severely damaged Dense white particles Many pinholes Illustrative example (Me-SiO 2, 135 C, 7 days)

10 Pore size distribution new membranes Permporometry with H 2 O Suitable for thin toplayers All membranes similar N 2 and mercury adsorption methods are for bulk material, not for thin layers! Normalised permeance Kelvin diameter (nm) Methylated Silica Bridged Silica Zirconia Titania 3 4

11 Pervaporation with zirconia membrane J H2 O (kg.m-2.h -2 ) Water content permeate (%) Time (days) 95 C, n-butanol/h 2 O (95/5%) Flux decline, but stable after 20 days Operating for at least 120 days

12 Hybrid membranes from bisfunctional silica precursors replacing Si O Si bonds by Si C C Si bonds (OC 2 H 5 ) 3 Si CH 2 CH 2 Si (OC 2 H 5 ) 3 (bis(triethoxysilyl)ethane, BTESE) Patented in collaboration with Univ. of Twente and Univ. of Amsterdam (Ashima Sah, Andre ten Elshof, Hessel Castricum, Marjo Mittelmeijer)

13 New membrane materials: Hybrid membranes BTESE Water, HNO 3, EtOH Reflux at 60ºC, 3 hrs Mix Coating and calcination Mix in EtOH MTES

14 Performance hybrid membranes (1) (150ºC) Water flux (kg/m 2 h) Time on stream (d) Watercontent in the permeate (%) Time on stream (d) Me-SiO 2 stability Feed: 5% H 2 O in BuOH T = 150 C

15 Hydrothermal stability at 150ºC Time 2y <2d 2w Si CH 3 -Si -Si-CH 2 CH 2 -Si

16 Performance hybrid membranes (2) (190ºC) Feed: 2,5% H 2 O in BuOH T = 190 C Water flux (kg/m 2 h) Time on stream (d) Waterconcentrate in the permeate (%)

17 Performance hybrid membranes (3) Water flux and conc. in perm in different solvents Water flux Water conc. in permeate Flux (g/m 2 h) MeOH EtOH PrOH BuOH MEK THF AcNi NMP Solvents Water in perm. (wt.%) Feed = 5 wt.% water in solvent

18 Origins of stability More stable bonds Higher crack propagation energy Lower surface diffusion coefficient Lower solubility

19 Hybrid membrane in the press Over 1700 hits on Chem Commun in February 2008 alone! NRC Chem Engineering MIT Tech Review Chem Processing Control & Automation Magazine Chem Tech C2W Tech Weekblad Petrochem

20 Next steps FOCUS: IMPLEMENTATION State of the art membrane Determine application window ph, H 2 O content, solvents Create consortium for commercialisation: end user(s), membrane producer(s), system integrator(s), supplier(s) enabling parts. Definition launching application(s). Further development: Reduce pore size: H 2 O-EtOH, and hydrogen separation Increase pore size: nanofiltration

21 Next step: 1m 2 field test

22 Acknowledgements The MST group at ECN Universities of Twente and Amsterdam: Hessel Castricum, Andre ten Elshof, Ashima Sah, Marjo Mittelmeijer- Hazeleger Financial support: STW, SenterNovem

23 Thank you for your attention For more information please contact me:

24 Chem. Commun., 2008, , DOI: /b718082a J. Mater. Chem., 2008, 18, , DOI: /b801972j J. Sol-Gel Sci Techn, 2008, DOI: /s z J. Mem. Sci, 2008, in preparation PLUS MORE TO COME! Patent: WO