Cellulose ultrafiltration membranes for cleaning solvent wastes of photolithography process


Creative Commons License

Savaş Alkan A., Çulfaz Emecen P. Z.

ICOM 12th International Congress on Membranes and Membrane Processes, London, Birleşik Krallık, 7 - 11 Aralık 2020, ss.1513

  • Yayın Türü: Bildiri / Özet Bildiri
  • Basıldığı Şehir: London
  • Basıldığı Ülke: Birleşik Krallık
  • Sayfa Sayıları: ss.1513
  • Orta Doğu Teknik Üniversitesi Adresli: Evet

Özet

Photolithography, the process used to transfer shapes and patterns on semiconductors, uses a number of solvents such as acetone, n-methyl pyrrolidone and propylene glycol monomethyl ether acetate (PGMEA), to clean surfaces and remove photoresists used throughout the patterning process. Cleaning these solvents for recycle in the photolithography process or for other purposes can have significant impact on the environmental burden of electronics manufacturing.

 

Cellulose is a naturally occurring polymer resistant to various solvents, including polar aprotics, which are the harshest for many synthetic membrane polymers, thus making cellulose membranes attractive for separations in solvent media.

 

In this study, cellulose acetate ultrafiltration membranes were prepared from solutions of cellulose acetate, dimethyl sulfoxide, acetone and polyethylene glycol (400 Da) (Table 1). Conversion to cellulose was carried out after fabrication by immersing the membranes in 0.05 M aqueous NaOH solution for 24 hours. In cases where annealing was done, cellulose acetate membranes were kept at 85°C for 3 hours before the alkaline hydrolysis step. Permeance of pure water (PWP) and PGMEA were measured in dead end cells and molecular weight cut-off (MWCO) was determined using PEG probes (400-20,000 Da) in water and Gel Permeation Chromatography. Rejection of the photoresist SU-8 was determined by measuring the SU-8 concentration via UV-Visible Spectrophotometry at a wavelength of 277 nm.

 

Table 1. Membrane fabrication conditions

Membrane Code

CA

DMSO

Acetone

PEG 400

Annealing

Alkaline hyrolysis

CA20

20

80

-

-

-

-

CA20P10

20

70

-

10

-

-

CA25P10

25

65

-

10

-

-

CA25P10A10

25

55

10

10

-

-

CA20-AH

20

80

-

-

-

+

CA20P10-AH

20

70

-

10

-

+

CA25P10-AH

25

65

-

10

-

+

CA25P10A10-AH

25

55

10

10

-

+

CA25P10-AN-AH

25

65

-

10

+

+

CA25P10A10-AN-AH

25

55

10

10

+

+

 

PWP and MWCO of membranes in Table 1 are shown in Figure 1(a) and 1(b), respectively. Adding PEG to the casting solution increased while increasing the CA concentration decreased the PWP. Alkaline hydrolysis in general increased the PWP, while the MWCO remained essentially unchanged around 10 kDa. Annealing the cellulose acetate membranes before hydrolysis decreased the MWCO to 3 kDa. The membrane CA25P10A10-AN-AH showed 85% rejection for uncrosslinked SU-8 resin, with a permeance of 0.2 L/hm2bar for the SU-8 developer solvent, PGMEA, used in photolithography.

 

Figure 1. Pure water permeance and Molecular Weight Cut-Off of the membranes

 

Acknowledgement

 

The study is funded by TUBITAK Project MAG218M509.

 

Photolithography, the process used to transfer shapes and patterns on semiconductors, uses a number of solvents such as acetone, n-methyl pyrrolidone and propylene glycol monomethyl ether acetate (PGMEA), to clean surfaces and remove photoresists used throughout the patterning process. Cleaning these solvents for recycle in the photolithography process or for other purposes can have significant impact on the environmental burden of electronics manufacturing.

 

Cellulose is a naturally occurring polymer resistant to various solvents, including polar aprotics, which are the harshest for many synthetic membrane polymers, thus making cellulose membranes attractive for separations in solvent media.

 

In this study, cellulose acetate ultrafiltration membranes were prepared from solutions of cellulose acetate, dimethyl sulfoxide, acetone and polyethylene glycol (400 Da) (Table 1). Conversion to cellulose was carried out after fabrication by immersing the membranes in 0.05 M aqueous NaOH solution for 24 hours. In cases where annealing was done, cellulose acetate membranes were kept at 85°C for 3 hours before the alkaline hydrolysis step. Permeance of pure water (PWP) and PGMEA were measured in dead end cells and molecular weight cut-off (MWCO) was determined using PEG probes (400-20,000 Da) in water and Gel Permeation Chromatography. Rejection of the photoresist SU-8 was determined by measuring the SU-8 concentration via UV-Visible Spectrophotometry at a wavelength of 277 nm.

 

Table 1. Membrane fabrication conditions

Membrane Code

CA

DMSO

Acetone

PEG 400

Annealing

Alkaline hyrolysis

CA20

20

80

-

-

-

-

CA20P10

20

70

-

10

-

-

CA25P10

25

65

-

10

-

-

CA25P10A10

25

55

10

10

-

-

CA20-AH

20

80

-

-

-

+

CA20P10-AH

20

70

-

10

-

+

CA25P10-AH

25

65

-

10

-

+

CA25P10A10-AH

25

55

10

10

-

+

CA25P10-AN-AH

25

65

-

10

+

+

CA25P10A10-AN-AH

25

55

10

10

+

+

 

PWP and MWCO of membranes in Table 1 are shown in Figure 1(a) and 1(b), respectively. Adding PEG to the casting solution increased while increasing the CA concentration decreased the PWP. Alkaline hydrolysis in general increased the PWP, while the MWCO remained essentially unchanged around 10 kDa. Annealing the cellulose acetate membranes before hydrolysis decreased the MWCO to 3 kDa. The membrane CA25P10A10-AN-AH showed 85% rejection for uncrosslinked SU-8 resin, with a permeance of 0.2 L/hm2bar for the SU-8 developer solvent, PGMEA, used in photolithography.

 

Figure 1. Pure water permeance and Molecular Weight Cut-Off of the membranes

 

Acknowledgement

 

The study is funded by TUBITAK Project MAG218M509.