- Email: [email protected]
Recovery of indium from TFT and CF glasses of LCD wastes using NaOH-enhanced sub-critical water
J. of Supercritical Fluids 104 (2015) 40–48
Contents lists available at ScienceDirect
The Journal of Supercritical Fluids journal homepage: www.elsevier.com/locate/supflu
Recovery of indium from TFT and CF glasses of LCD wastes using NaOH-enhanced sub-critical water Hiroyuki Yoshida a,c,∗ , Shamsul Izhar a,c , Eiichiro Nishio b , Yasuhiko Utsumi b , Nobuaki Kakimori b , Feridoun Salak Asghari a a
Research Organization for the 21st Century, Osaka Prefecture University, 1-1 Gakuencho, Nakaku, Sakai-shi, Osaka 599-8570, Japan Environment Technology Development Center, Environment and Safety Division, Sharp Corporation, 1 Takumicho, Sakaiku, Sakai, Osaka 590-8522, Japan c Department of Chemical and Environmental Engineering, Faculty of Engineering, Universiti Putra Malaysia, 43400 UPM Serdang, Selangor, Malaysia b
a r t i c l e
i n f o
Article history: Received 2 April 2015 Received in revised form 18 May 2015 Accepted 18 May 2015 Available online 19 May 2015 Keywords: LCD panel Indium recovery Sub-critical water in the presence of NaOH TFT glass CF glass
a b s t r a c t Recently, we reported that 83% indium was recovered from CF glass and 10% from TFT glass of LCD panel wastes using sub-critical water at 360 ◦ C. In the present work, in order to increase the recovery of indium in the form of indium tin oxide (ITO) and to reduce the reaction temperature, the effect of basic materials and their concentrations in sub-critical water on its recovery were tested. The basic materials were NaOH, KOH, Na2 CO3 , diethyl amine, Ca(OH)2 and NH3 . NaOH showed the largest recovery of indium from both CF and TFT glasses and reduced the reaction temperature. Treatment for only 5 min resulted in an outstanding indium recovery of 95% at 220 ◦ C from TFT glass and 99% at 160 ◦ C from CF glass. After the treatment, both TFT and CF glasses became transparent. ITO did not dissolve in the liquid-phase but remained in the solid organic multi-layers that were separated from the TFT and CF glasses by the subcritical water reaction. These organic multi-layers were readily recovered by filtration. Since no indium dissolved in the liquid-phase, this recovery method is superior to others such as acid dissolution and ionexchange. The optimum requirement of NaOH concentration and reaction time for indium recovery was 0.1 N and 5 min, respectively. The present result showed a significant improvement in indium recovery compared to our previous result without NaOH. Thus, sub-critical water in the presence of NaOH is highly feasible for the recovery of indium from TFT and CF glasses. © 2015 Elsevier B.V. All rights reserved.
1. Introduction A thin film transistor type liquid crystal display (TFT-LCD) panel generally consists of layers of polymer, thin-film transistor (TFT), color filter (CF) substrates, and liquid crystals. The TFT glass and CF glass contain complex layers of electrodes, multiple organic films, terminals, and semi-conductors. The key materials for manufacturing TFT and CF glasses are the rare earth indium and valuable glass substrates with high heat resistance and high mechanical strength. The indium is employed in the form of ITO (indium tin oxide) [1] and is used for transparent electrodes due to its low resistance and patterning capability. However, indium is limited in production and its market price has increased [2]. Furthermore, an increasing quantity of LCD wastes is expected in a few years, thus it is essential to
∗ Corresponding author at: Department of Chemical and Environmental Engineering, Faculty of Engineering, Universiti Putra Malaysia, 43400 UPM Serdang, Selangor, Malaysia. E-mail address: [email protected] (H. Yoshida). http://dx.doi.org/10.1016/j.supflu.2015.05.016 0896-8446/© 2015 Elsevier B.V. All rights reserved.
recover precious resources such as indium from LCD panel wastes [3]. Efforts to recover valuable materials, especially indium from LCD wastes and ITO targets, have been investigated by ionexchange [4] and acid leaching [5,6]. Solvent extraction [7,8], hydrometallurgical [9] and chlorination [10,11] methods have also been reported to recover ITO from etching solutions or target. However, due to complications in the above processes, a simpler treatment method by means of sub-critical water (hereafter called sub-CW) technology was initiated by our group [12]. This technique is based on the use of water at temperatures between 100 ◦ C and 374 ◦ C, and at pressures high enough to maintain the liquid state. The magnitude of the ionic product of water increases three orders at around 250 ◦ C compared to room temperature [13]. This property is advantageous to facilitate hydrolysis and the decomposition of organic compounds. In addition, the dielectric constant of water decreases, thereby lowering its polarity. Thus, there are many reports on sub-CW for waste treatment and utilization [14–18]. In our previous paper [12] using sub-critical water at 360 ◦ C and a reaction time of 5 min, 83% indium recovery from CF glass was
H. Yoshida et al. / J. of Supercritical Fluids 104 (2015) 40–48
41
reported. Transparent glass was also recovered at the same time from CF glass. However, in the TFT glass, only about 10% of the indium was recovered. This low yield was due to the insufficient decomposition of organic matter from the TFT glass. The removal of the organic multilayer and the hydrolyzed layer required more aggressive decomposition reaction conditions than those employed in the previous study. The objectives of the present work are to obtain a higher indium recovery using sub-CW treatment and to reduce the reaction temperature. Suyama et al. found that basic materials, like alkylamines and organic compounds like long alkyl chained alcohol, used as additives in sub-CW are efficient to decompose polyesters and polystyrene [19–21]. In the present work, the possibility of using several basic inorganic hydroxides Na2 CO3 , ammonia and amine compounds in sub-CW is tested to increase the yield of indium and reduce the reaction temperature. At the same time, the recovery of clean glass substrate from the remaining treated LCD is also investigated.
2. Materials and experimental procedure 2.1. Preparation of LCD glass sample The materials and experimental procedure carried out in the present study were the same as those reported in our previous work [12]. In brief, the LCD panel used in this work was a 40 inch TFT type LCD panel, composed of TFT glass and CF glass. The polarizing films were removed from the panel. Then the panel was separated into CF glass and TFT array using a glass cutter (Toshin-Riko). Liquid crystals on both glass surfaces were washed away using acetone. Thereafter, both glasses were broken into about 5–10 mm sizes, small enough to fit and be in motion in the reactor. From here onwards, CF glass and TFT glass denote glass specimens without the liquid crystal layer and polarizing films.
2.2. Sub-CW water treatment The reaction field tested in this work was ultrapure water (MilliQ < 0.05 S cm−1 ) and aqueous solutions of NaOH, KOH, Na2 CO3 , Ca(OH)2 , ammonia and diethylamine (DEA), respectively. All reagents were analytical grade. Each aqueous solution was made using MilliQ water. The batch reactors used involved a stainless steel pipe (SUS316, 16 mm id, 150 mm length, 30 × 10−6 m3 ) and stainless steel tube end-locks (Swagelok). First, one side of the pipe was tightly sealed using the end-lock. After pouring 20 cm3 of each solution together with 6 g of either CF glass or TFT glass into the reactor, the air gap was replaced with argon gas. Thereafter, the reactor was sealed with the Swagelok cap and was then weighed. It was then immersed quickly in a preheated molten salt bath (Celsius, Thomas Kagaku Co. Ltd., Tokyo) containing a mixture of potassium nitrate and sodium nitrate with a stable temperature, and shaken. After a period of time, the reaction tube was taken out of the salt bath and immediately quenched in a tap water bath at room temperature. The temperature inside the reactor instantly increased until it equals the salt bath temperature [22]. The reaction time was defined as the period of time from immersing the reactor in the salt bath until transferring it into the tap water bath. The contents were recovered from the reactor and their physical changes were observed. The solid residues were separated from the aqueous phase by filtration through a 1.0 m pore-sized membrane filter (Advantec, cellulose acetate). The reactor tube and cap were rinsed with MilliQ water and the water was filtered together with the aqueous phase.
Fig. 1. Effect of NaOH, KOH, Na2 CO3 , DEA, Ca(OH)2 and NH3 in sub-CW on indium recovery. All additive concentrations were 0.1 N. Reaction temperatures and time was 260 ◦ C and 5 min, respectively.
2.3. Measurement and analysis After sub-CW treatment, the metal content in each medium was determined as follows: the indium on the glass and on the membrane filter was extracted using a known amount of 7% HCl aq. solution, respectively, according to Eq. (1): In2 O3 + 6HCl → 2InCl3 + 3H2 O(1) Then they were sonicated for 1 h at 50 ◦ C. The indium concentration in the HCl aq. solution was measured using ICP (Shimadzu, ICPE-9000) and the amounts of indium in the fresh glass (W0 ) and in the remaining glass substrate (Wg ), filter (Wf ) and liquid-phase (Wl ) were determined. The material balance of indium before and after the reaction was confirmed. The recovery of indium (in ) was calculated using Eq. (2): in =
Wf × 100 W0
(2)
Subsequent to the indium analysis, Sn, Si, and Fe concentrations were also determined from each sample by ICP. The total organic carbon content in the liquid-phase was measured using a TOC analyzer (Shimadzu, TOC-Vcph/cpn). The metallurgical microstructure images of the glass surface were obtained using an optical fluorescence microscope (BX51 Olympus Corp.). 3. Results and discussion 3.1. Effect of basic materials in sub-CW Our previous paper [12] showed that the maximum recovery of indium from TFT glass by sub-CW was only 10%. To increase the recovery, the effect of basic materials in sub-CW on indium recovery was tested. Fig. 1 shows the effect of the basic material in sub-CW on the recovery of indium from TFT glass at 260 ◦ C for 5 min. Without basic materials (water only), only 5% of the indium was recovered. However, when sub-CW contained 0.1 N NaOH, KOH, Na2 CO3 and Ca(OH)2 , respectively, the indium recovery increased to 90%, 85%, 60%, and 25%, respectively. The effects of 0.1 N NH3 and DEA were also investigated and resulted in an indium recovery of 2% and 8%, respectively. Sub-CW treatment with the presence of alkali increased the indium recovery. Alkali which contained OH− demonstrated a higher indium recovery than without
42
H. Yoshida et al. / J. of Supercritical Fluids 104 (2015) 40–48
the hydroxide ion. Suyama et al. [19–21] reported that amines were effective to improve hydrolysis. In this study, KOH and NaOH showed superior indium recovery compared to other additives because they are strong bases due to their smaller pKb (0.5 and 0.2, respectively). In addition, Ca(OH)2 , DEA and NH3 in which their pKb are 2.4, 3.2, and 4.7, respectively, are weaker base and thus resulted in less indium recovery. Consequently, the result showed a good relation between the pKb and the indium recovery. Since the addition of NaOH produced the best indium recovery and the price of NaOH is cheaper than KOH, only NaOH was used in further studies in this work. 3.2. Sub-CW treatment of TFT glass Fig. S1 (in Supplemental information) demonstrates photographs of the TFT glass after treatment of sub-CW with 0.1 N NaOH for 5 min at reaction temperatures between 120 and 360 ◦ C. Below 160 ◦ C, no change of color was observed in the aqueous solution, but at 180 ◦ C it became slightly brown. At 200 ◦ C, brownish floating materials, which were possibly the organic film, were observed in large amounts. At around 240 ◦ C, the color of the solution transformed from light brown to intensely brown. This is because the organic film on the surface of the TFT glass was hydrolyzed due to the sub-CW reaction and then changed to some substances that were water soluble. The aqueous solution was then filtered through the membrane filter. Fig. S2 (in Supplemental information) shows the photographs of the filters. At 200 ◦ C and 220 ◦ C, the brownish water-insoluble materials were observed on the filter. They were the organic multilayer films wrapped together with ITO exfoliated from TFT glass. This phenomenon suggested that the leaching of indium did not occur because the solid film layers did not dissolve in the aqueous solution. The concentration of indium in the aqueous phase is given below. However, above 240 ◦ C, the organic film became smaller, and started to shrink above 260 ◦ C due to the decomposition of the organic material. 3.3. Sub-CW treatment of CF glass Fig. S3 (in Supplemental information) shows photographs of the CF glass after treatment of sub-CW with 0.1 N NaOH between 100 and 360 ◦ C for 5 min. At 100 ◦ C, the aqueous solution was clear and no physical change was observed. However at 120 ◦ C, the color of the solution was yellow, before becoming extremely yellow at 140 ◦ C. Furthermore, dye pigments consisting of red, green, and blue were observed floating in large amounts in the solution. At temperatures above 160 ◦ C, the solution changed to dark blue before becoming dark brown at 320 and 360 ◦ C. After sub-CW treatment, the aqueous solution was filtered through the membrane filter. Photographs of the membrane filter surface after the filtration are shown in Fig. S4 (in Supplemental information). At 100 ◦ C, nothing was detected on the filter. The release of substantial amounts of water-insoluble green pigments were observed at 120 ◦ C, then red pigments at 140 ◦ C, and blue pigments at 160 ◦ C. The color became dark above 180 ◦ C. These phenomena showed that the exfoliation of the color pigments occurred in subsequent steps from green, red, and blue. As a consequence, these filtered water-insoluble CF materials were present due to the decomposition of the organic films on the CF glass surface caused by the hydrolysis reaction in sub-CW with 0.1 N NaOH. 3.4. Recovery of indium from TFT and CF glasses In order to evaluate the amount of indium recovery by subCW reaction, the amounts of Wg , Wf and Wl were determined and
Fig. 2. Effect of reaction temperature on presence of indium in each phase after treatment of sub-CW with 0.1 N NaOH. (a) TFT glass and (b) CF glass. Reaction time of 5 min.
expressed in milligrams of indium per kilogram of glass substrate (mg/kg-GL). Fig. 2(a) shows the results after treatment of TFT glass by sub-CW with 0.1 N NaOH. An excellent mass balance of indium quantity that was distributed in each phase was obtained. Indium was not detected at any temperatures in the liquid-phase. Treatment below 160 ◦ C showed no decrease of Wg , denoting that no indium was removed from TFT glass. However, above 160 ◦ C, Wg decreased sharply from 290 mg/kg-GL to 16 mg/kg-GL (200 ◦ C). This suggests that a large amount of indium was recovered from the TFT glass surface. Wg remained below 10 mg/kg-GL at 220–320 ◦ C but increased to 70 mg/kg-GL at 360 ◦ C. Meanwhile, Wf showed a trend opposite to Wg . Wf was zero below 160 ◦ C, but then increased sharply at 180–200 ◦ C. At 220 ◦ C, Wf was the highest with almost 290 mg/kg-GL. At 230–320 ◦ C, Wf was a little smaller than the highest value. However, Wf decreased when the reaction temperature was above 340 ◦ C, indicating that some indium remained on the TFT glass. Fig. 2(b) shows the results in CF glass. Indium was not detected in the liquid-phase. Even at 100 ◦ C, Wg reduced to 162 mg/kgGL. At 140 ◦ C, Wg significantly decreased to below 5 mg/kg-GL. It remained below 5 mg/kg-GL between 140 and 320 ◦ C. However at 360 ◦ C, Wg slightly increased to 40 mg/kg-GL. Wf showed the opposite trend to Wg . At 100 ◦ C, Wf was 170 mg/kg-GL. It increased to 350 mg/kg-GL at 140 ◦ C. At reaction temperatures between 140 ◦ C and 320 ◦ C, Wf remained high (between 340 and 360 mg/kg-GL) but slightly dropped to 300 mg/kg-GL at 360 ◦ C. Thus, treatment of sub-CW with 0.1 N NaOH between 140 and 320 ◦ C resulted in a complete exfoliation of CF materials, including indium, from the CF glass and they could be completely recovered by the membrane filter.
H. Yoshida et al. / J. of Supercritical Fluids 104 (2015) 40–48
43
3.5. Optic microscope images of TFT and CF glasses
Fig. 3. Effect of NaOH on the indium recovery after sub-CW reaction over (a) TFT and (b) CF glasses for 5 min at various temperatures. The solid and dotted lines indicate sub-CW reaction with 0.1 N NaOH and without NaOH [12], respectively.
The reason why Wg increased and Wf decreased in both TFT and CF glasses when the reaction temperature was higher than 340 ◦ C will be discussed in a later section. Fig. 3 illustrates the effect of reaction temperature on the indium recovery (in ) after treatment of sub-CW with and without 0.1 N NaOH. Fig. 3a shows the results in TFT glass. The dotted line shows the results without NaOH and in was only less than 10% [12]. However, when 0.1 N NaOH existed in sub-CW and the reaction temperature was higher than 160 ◦ C, in increased to 95% at 220 ◦ C. The in was around 93% between 220 and 320 ◦ C. This shows that the treatment of sub-CW with 0.1 N NaOH was far superior compared to that without NaOH. TFT substrates are usually coated with a thin (∼80–100 nm) alignment layer, insulating the optomer layer and protecting the passivation layers (see Fig. S1 in [12]). The optomer layer is an organic mixture containing mainly diethylene glycol methyl ethyl ether (MEC) and acrylic resin. Meanwhile, the passivation layer covering the drain and source terminals are made of SiNx . This suggested that the decomposition of the optomer and passivation layers on the TFT glass by sub-CW was enhanced drastically by the presence of NaOH. In the CF glass in Fig. 3(b), the effect of NaOH was also clear. When 0.1 N NaOH existed in sub-CW, in was 45% even at 100 ◦ C. This means that the hydrolysis reaction was active and had already taken place even at the water boiling point. in increased to 99% at 160 ◦ C. Between 160 and 320 ◦ C, in was somewhere between 90 and 99%. The dotted line shows the result without NaOH [12], where it was much smaller than that with 0.1 N NaOH. CF glass consists of 2–3 m thin color pigments, an overcoat layer and a 1 m thin black matrix [23,24]. These organic substances may be more susceptible to attacks from OH− in sub-CW. Thus exfoliation of these multiple organic layers occurred even at 100 ◦ C, which was lower than that in TFT glass.
Fig. 4 shows TFT glass surfaces observed using an optical microscope before and after treatment of sub-CW with 0.1 N NaOH. No difference was observed for (a) before reaction and (b) treated at 120 ◦ C. In (c) treated at 160 ◦ C, it became slightly distorted and showed early signs of exfoliations of the organic multilayer film. At 200 ◦ C (d), the glass surface became partially visible and this showed that the organic multilayer film was stripping away. The materials on the membrane filter were therefore attributed to the presence of the exfoliated organic multilayer films (Fig. S2). At 220 ◦ C (e), a clear transistor surface was observed with the gate and source bus-line terminals still remaining in an intact condition, and the whole organic multilayer film was exfoliated as evidenced from the filtered material (Fig. S2). This led to the extremely high indium recovery at 220–320 ◦ C as observed in Figs. 2 and 3. At 360 ◦ C (f), the source and gate bus-lines were slightly detached and grayish matter appeared on the surface as if a layer was formed. Fig. 5 shows optical microscope images of the CF glass before reaction (a) and after treatment of sub-CW with 0.1 N NaOH. At 100 ◦ C (b), the pigment colors became paler compared to (a). At 120 ◦ C (c), 140 ◦ C (d) and 160 ◦ C (e), exfoliations of green pigment followed by red and blue pigments were observed. This agrees with the results in Fig. S4, where the filtered material was green, combined red–green and combined blue–red–green at 120 ◦ C, 140 ◦ C, and 160 ◦ C, respectively. The black matrix was slightly visible at 160 ◦ C (e). At 280 ◦ C (f), the glass was transparent, without the color pigments and black matrix. This indicates that the alignment film, overcoat layer, ITO and color filters were totally exfoliated from the CF glass. This was clear in Fig. S4, where the filtered material was completely dark, with a tremendously high indium recovery between 180 and 280 ◦ C. 3.6. Mechanism of exfoliation of multi-layered organic membranes from LCD The exfoliation mechanisms in sub-CW with 0.1 N NaOH are further discussed in Fig. 6. In the TFT glass (Fig. 6(a)), the multi-layered organic membranes that consist of alignment, optomer and passivation layers were not removed below 160 ◦ C, but to some extent were removed at 200 ◦ C. Since ITO was located below the alignment film, the ITO exfoliated from TFT glass during the sub-CW treatment was in a state which was sandwiched between the organic films. Thus, the indium recovery at 200 ◦ C is ascribed to the partial exfoliation of the organic multilayer film. However at 220 ◦ C, a complete exfoliation of the layers occurred, thus the whole of the layers including ITO was removed because of the strong hydrolysis of sub-CW with 0.1 N NaOH. This strong hydrolysis is explained by the maximum ionic product of water at around 250 ◦ C as shown in Fig. S5 (in Supplemental information). Fig. 6(b) illustrates the mechanism of exfoliation of the multiple layers from CF glass. The pale color of the microscopic figure could be due to the partial removal of the alignment layer and ITO at 100 ◦ C, where about 50% of them may have been removed. Thus, indium recovery was about 50%. At 120, 140, and 160 ◦ C, the color pigments of red, green, and blue, respectively, were removed in turn from the surface but the black matrix still remained on it. At 280 ◦ C, all of the alignment and ITO layers, the color pigments and the black matrix were exfoliated from the glass. This was confirmed by the transparent glass photo from the microscopic image in Fig. 5(f) and the color of the filtered material in Fig. S4. Fig. 2 showed that Wg at 360 ◦ C was larger than those at lower temperature. This phenomenon is related to the fact that the hydrolysis activity shows a maximum at around 250 ◦ C and gets weaker when the reaction temperature is higher than 300 ◦ C (Fig.
44
H. Yoshida et al. / J. of Supercritical Fluids 104 (2015) 40–48
Fig. 4. Microscope images of (a) untreated TFT glass, and the TFT glass treated by sub-CW with 0.1 N NaOH for 5 min at (b) 120 ◦ C, (c) 160 ◦ C, (d) 200 ◦ C, (e) 220 ◦ C, and (f) 360 ◦ C, respectively.
Fig. 5. Microscope images of the (a) untreated CF glass, and CF glass treated by sub-CW with 0.1N NaOH for 5 min at (b) 100 ◦ C, (c) 120 ◦ C, (d) 140 ◦ C, (e) 160 ◦ C, and (f) 280 ◦ C, respectively.
S5). As at 360 ◦ C, the hydrolytic strength becomes weaker than that at 250 ◦ C, the mechanisms of exfoliation in TFT and CF glasses are explained as Fig. 6(A) and (B) at 360, respectively. 3.7. TOC and Si concentrations of the liquid-phase TOC was measured to make clear how much water-soluble organic material in the liquid-phase was produced by sub-CW with 0.1 N NaOH. Fig. 7 shows the effect of the reaction temperature
on TOC in the liquid-phase. TOC concentrations from both TFT and CF glasses clearly are much higher than those without NaOH [12]. They escalated with increasing temperature. In TFT glass, the TOC concentration was zero at T < 180 ◦ C. TOC increased sharply to 1.34 mg/g-GL in 180 < T <260 ◦ C. At 260 < T < 360 ◦ C, TOC was about 1.34 mg/g-GL. On the other hand, in the CF glass, TOC concentration was 0.2 mg/g-GL even at 100 ◦ C and increased sharply to 1.13 mg/gGL at 180 ◦ C, then gradually increased and was close to that of TFT glass at T > 260 ◦ C.
H. Yoshida et al. / J. of Supercritical Fluids 104 (2015) 40–48
45
Fig. 6. Mechanism of exfoliation of organic multiple layer from TFT and CF glasses during treatment of sub-CW with 0.1 N NaOH at various temperatures (Reaction time: 5 min).
The TOC concentration in CF is higher than that in TFT in T < 250 ◦ C. This is evidence that the organic layers in CF and TFT glasses are different, and thus would require different reaction temperatures for exfoliation from the glasses. To ascertain the exfoliation of the passivation layer that consisted of SiNx , the silica concentration in the liquid-phase was determined by ICP by a similar extraction method to when determining the indium concentration. Fig. 8 shows Si concentration in each phase after TFT and CF glasses were treated by sub-CW with 0.1 N NaOH. In the TFT glass, Si concentration was less than 500 mg/kg-GL in all phases when treated below 190 ◦ C. However,
Si concentration in the liquid-phase increased sharply to about 3800 mg/kg-GL between 280 and 320 ◦ C. This suggested that SiNx from the passivation layer has dissolved into the liquid-phase as a result of the exfoliation of the organic multi-layers with ITO. On the other hand, the increase of Si concentration was not observed in the CF glass. This is reasonable since CF glass does not contain any passivation layer. Above 300 ◦ C, the silica concentrations on the glass increased in both TFT and CF glasses. This is probably due to the corrosion of the substrate glass in NaOH environment at a very high reaction temperature causing the formation of sodium silicate.
46
H. Yoshida et al. / J. of Supercritical Fluids 104 (2015) 40–48
Fig. 7. Effect of temperature on total organic carbon (TOC) concentrations of liquid phase after treatment of TFT and CF glasses by sub-CW with 0.1 N NaOH (reaction time: 5 min). Dotted line indicates the sub-CW treatment without NaOH [12].
Fig. 9. NaOH concentration of the resultant solution as a function of reaction temperature. Reaction time: 5 min.
Fig. 8. Presence of silica in each phase during sub-CW reaction with 0.1 N NaOH and reaction time of 5 min as a function of reaction temperatures. (a) TFT glass and (b) CF glass.
Fig. 9 illustrates the NaOH concentration of the resultant solution that was measured by titration. The NaOH concentration from TFT glass was 90 mM from 120 ◦ C until about 270 ◦ C before it dropped sharply to 20 mM (360 ◦ C). The NaOH concentration from CF glass dropped steadily from 90 mM (100 ◦ C) to 85 mM (270 ◦ C) before it sharply dropped to 20 mM (360 ◦ C). This supports the evidence that SiNx from the passivation layer has dissolved into the liquid-phase as a result of the exfoliation of the organic multilayers. This is followed by corrosion of the substrate glass in NaOH environment above 270 ◦ C causing the formation of sodium silicate. 3.8. Effect of NaOH concentration and reaction time. Fig. 10 shows the effect of the concentration of NaOH on indium recovery. The reaction time was 5 min while the reaction temperatures were taken from the optimum reaction temperatures for TFT and CF glasses which were 220 and 160 ◦ C, respectively. No increase in in was observed between 0.003 and 0.01 N in both TFT and CF glasses. When the concentration was above 0.01 N, in increased sharply and was above 95% at 0.1 N. in of both glasses remained almost constant above 0.1 N. This clearly shows that 0.1 N NaOH is the optimum for indium recovery from both TFT and CF glasses.
Fig. 10. Effect of NaOH concentration on indium recovery over (a) TFT glass at reaction temperature of 220 ◦ C and (b) CF glass at 160 ◦ C (Reaction time: 5 min).
The optimum reaction time was investigated in 0.1 N NaOH for TFT and CF glasses at 220 ◦ C and 160 ◦ C, respectively. The results are given in Fig. 11. in in TFT and CF glasses increased sharply to about 80% in only 2 min. At 5 min, in reached 95% for TFT glass and 99% for CF glass. No further increase was observed for reaction times longer than 5 min. This indicates that a reaction time of 5 min is sufficient for both TFT and CF glasses.
H. Yoshida et al. / J. of Supercritical Fluids 104 (2015) 40–48
47
was recovered just by filtration, which is a much simpler recovery process than acid dissolution, chlorination and ion-exchange methods. The amount of indium recovered from CF glass and TFT glass was dependent on the reaction temperature. The optimum temperatures for TFT and CF glasses were 220 ◦ C and 160 ◦ C, respectively. The indium recoveries from TFT and CF glasses were 95% and 99%, respectively. The minimum requirements for indium recovery were 0.1 N for NaOH concentration and 5 min of reaction time. The dissolution of the Fe ion from the reactor wall was small, demonstrating the vessel wall can withstand sub-critical conditions with 0.1 N NaOH. With the low cost and low corrosion effect of NaOH, the treatment of sub-CW with NaOH has high potential to be applied in the recovery of indium from LCD panel wastes. Appendix A. Supplementary data Supplementary data associated with this article can be found, in the online version, at http://dx.doi.org/10.1016/j.supflu.2015.05. 016 References
Fig. 11. Effect of reaction time on indium recovery over (a) TFT glass and (b) CF glass at reaction temperature of 220 ◦ C and 160 ◦ C, respectively.
3.9. Corrosion of the reactor wall Concerning its commercial application, the sub-CW apparatus should be used for a long period. The corrosion of the surface of the reactor in sub-CW in the presence of 0.1 N NaOH was tested. Fig. S6 shows the dissolution of the Fe ion from the internal surface of the batch reactor (SUS316) at 260 ◦ C. During the first 50 min, the Fe ion increased sharply from 0 to 0.28 mg Fe/20 mL NaOH. However beyond 50 min, the Fe ion increased slightly to 0.35 mg Fe/20 mL NaOH. This showed that the Fe dissolution rate of the vessel wall was tremendously slow. Assuming a homogeneous dissolution of the wall, it is predicted that the wall thickness lessens by a mere 2 m a year. Thus, due to this small effect, it is considered that vessel walls can withstand continuous and extensive operation when treated by sub-CW in the presence of 0.1 N NaOH. 4. Conclusions Indium in the form of ITO was recovered from TFT and CF glasses by treatment of sub-CW in additives. In the presence of NaOH, KOH, Na2 CO3 , Ca(OH)3 , and NH3 , respectively, NaOH gave the highest recovery of indium. Sub-CW treatment in 0.1 N NaOH resulted in a near complete recovery of indium from TFT and CF glasses in just a short period of 5 min. This result was a remarkable progress from our previous study using only water [12]. ITO was recovered from the glass together with the organic multi-layers that did not dissolve in the liquid-phase. This is very beneficial because indium
[1] A.M. Alfantazi, R.R. Moskalyk, Processing of indium: a review, Min. Eng. 16 (2003) 687–694. [2] G. Phipps, C. Mikolajczak, T. Guckes, Indium and Gallium: long-term supply, Renew. Energy Focus 9 (2008) 56–59. [3] S. Stalhofer, M. Spitzbart, K. Maurer, Recycling of LCD screens in Europe – State of the art and challenges, in: J. Hesselbach, C. Herrmann (Eds.), Globalized Solutions for Sustainability in Manufacturing, Springer-Verlag Berlin Heidelberg, 2011, pp. 454–458. [4] M. Tsujiguchi, Design for innovative value towards sustainable society, in: M. Matsumoto, Y. Umeda, K. Masui, S. Fukushige (Eds.), Proceedings of EcoDesign 2011: 7th International Symposium on Environmentally Conscious Design and Inverse Manufacturing, Springer, Dordrecht, 2012, pp. 743–746. [5] J. Li, S. Gao, H. Duan, L. Liu, Recovery of valuable materials from waste liquid crystal display panel, Waste Manage. 29 (2009) 2033–2039. [6] Y. Li, Z. Liu, Q. Li, Z. Liu, L. Zeng, Recovery of indium from used indium–tin oxide (ITO) targets, Hydrometallurgy 105 (2011) 207–212. [7] M.C.B. Fortes, A.H. Martins, J.S. Benedetto, Indium recovery from acidic aqueous solutions by solvent extraction with D2EHPA: a statistical approach to the experimental design, Sharp Tech. J. 92 (2005) 17–22. [8] T. Honma, T. Muratani, Material collection from liquid crystal display wasted panels, Sharp Tech. J. 92 (2005) 17–22. [9] S.J. Hsieh, C.C. Chen, W.C. Say, Process for recovery of indium from ITO scraps and metallurgic microstructures, Mat. Sci. Eng. B 158 (2009) 82–87. [10] K.S. Park, W. Sato, G. Grause, T. Kameda, T. Yoshioka, Recovery of indium from In2 O3 and liquid crystal display powder via a chloride volatilization process using polyvinyl chloride, Thermochim. Acta 493 (2009) 105–108. [11] K. Takahashi, A. Sasaki, G. Dodbiba, J. Sadaki, N. Sato, T. Fujita, Recovering Indium from the liquid crystal display of discarded cellular phones by means of chloride-induced vaporization at relatively low temperature, Metall. Mater. Trans. A 40 (2009) 891–900. [12] H. Yoshida, S. Izhar, E. Nishio, Y. Utsumi, N. Kakimori, S.A. Feridoun, Recovery of indium from TFT and CF glasses in LCD panel wastes using sub-critical water, Solar Energy Mater. Solar Cells 125 (2014) 14–19. [13] W. Wagner, H.-J. Kretzschmar, International Steam Tables: Properties of Water and Steam, 2nd ed., Springer-Verlag Berlin Heidelberg, 2008, pp. 322. [14] H. Yoshida, H. Tokumoto, K. Ishii, R. Ishii, Efficient, high-speed methane fermentation for sewage sludge using subcritical water hydrolysis as pretreatment, Bioresour. Technol. 100 (2009) 2933–2939. [15] O. Pourali, F.S. Asghari, H. Yoshida, Production of phenolic compounds from rice bran biomass under subcritical water conditions, Chem. Eng. J. 160 (2010) 259–266. [16] O. Tavakoli, H. Yoshida, Application of sub-critical water technology for recovery of heavy metal ions from the wastes of Japanese scallop Patinopecten Yessoensis, Ind. Eng. Chem. Res. 45 (2006) 5675–5680. [17] W. Abdelmoez, H. Yoshida, Kinetics and mechanism of the synthesis of a novel protein-based plastic using subcritical water, Biotechnol. Prog. 24 (2008) 466–475. [18] W. Abdelmoez, H. Yoshida, Mechanical and thermal properties of a novel protein-based plastic synthesized using subcritical water technology, Macromolecules 40 (2007) 9371–9377. [19] K. Suyama, M. Kubota, M. Shirai, H. Yoshida, Effect of alcohols on the degradation of crosslinked unsaturated polyester in sub-critical water, Polym. Degrad. Stab. 91 (2006) 983–986. [20] K. Suyama, M. Kubota, M. Shirai, H. Yoshida, Degradation of crosslinked unsaturated polyesters in sub-critical water, Polym. Degrad. Stab. 92 (2007) 317–322.
48
H. Yoshida et al. / J. of Supercritical Fluids 104 (2015) 40–48
[21] K. Suyama, M. Kubota, M. Shirai, H. Yoshida, Chemical recycling of networked polystyrene derivatives using subcritical water in the presence of an aminoalcohol, Polym. Degrad. Stab. 95 (2010) 1588–1592. [22] W. Abdelmoez, H. Yoshida, Simulation of fast reactions in batch reactors under sub-critical water conditions, AICHe J. 52 (2006) 3600–3611.
[23] T.M. Lee, Y.J. Choi, S.Y. Nam, C.W. You, D.Y. Na, H.C. Choi, D.Y. Shin, K.Y. Kim, K.I. Jung, Color filter patterned by screen printing, Thin Solid Films 516 (2008) 7875–7880. [24] P.M. Harrison, J. Wendland, M. Henry, Innovative laser patterning of black,atrix for LCD manufacture, Soc. Infor. Display (SID) Symp. Digest 39 (736) (2008).
Contents lists available at ScienceDirect
The Journal of Supercritical Fluids journal homepage: www.elsevier.com/locate/supflu
Recovery of indium from TFT and CF glasses of LCD wastes using NaOH-enhanced sub-critical water Hiroyuki Yoshida a,c,∗ , Shamsul Izhar a,c , Eiichiro Nishio b , Yasuhiko Utsumi b , Nobuaki Kakimori b , Feridoun Salak Asghari a a
Research Organization for the 21st Century, Osaka Prefecture University, 1-1 Gakuencho, Nakaku, Sakai-shi, Osaka 599-8570, Japan Environment Technology Development Center, Environment and Safety Division, Sharp Corporation, 1 Takumicho, Sakaiku, Sakai, Osaka 590-8522, Japan c Department of Chemical and Environmental Engineering, Faculty of Engineering, Universiti Putra Malaysia, 43400 UPM Serdang, Selangor, Malaysia b
a r t i c l e
i n f o
Article history: Received 2 April 2015 Received in revised form 18 May 2015 Accepted 18 May 2015 Available online 19 May 2015 Keywords: LCD panel Indium recovery Sub-critical water in the presence of NaOH TFT glass CF glass
a b s t r a c t Recently, we reported that 83% indium was recovered from CF glass and 10% from TFT glass of LCD panel wastes using sub-critical water at 360 ◦ C. In the present work, in order to increase the recovery of indium in the form of indium tin oxide (ITO) and to reduce the reaction temperature, the effect of basic materials and their concentrations in sub-critical water on its recovery were tested. The basic materials were NaOH, KOH, Na2 CO3 , diethyl amine, Ca(OH)2 and NH3 . NaOH showed the largest recovery of indium from both CF and TFT glasses and reduced the reaction temperature. Treatment for only 5 min resulted in an outstanding indium recovery of 95% at 220 ◦ C from TFT glass and 99% at 160 ◦ C from CF glass. After the treatment, both TFT and CF glasses became transparent. ITO did not dissolve in the liquid-phase but remained in the solid organic multi-layers that were separated from the TFT and CF glasses by the subcritical water reaction. These organic multi-layers were readily recovered by filtration. Since no indium dissolved in the liquid-phase, this recovery method is superior to others such as acid dissolution and ionexchange. The optimum requirement of NaOH concentration and reaction time for indium recovery was 0.1 N and 5 min, respectively. The present result showed a significant improvement in indium recovery compared to our previous result without NaOH. Thus, sub-critical water in the presence of NaOH is highly feasible for the recovery of indium from TFT and CF glasses. © 2015 Elsevier B.V. All rights reserved.
1. Introduction A thin film transistor type liquid crystal display (TFT-LCD) panel generally consists of layers of polymer, thin-film transistor (TFT), color filter (CF) substrates, and liquid crystals. The TFT glass and CF glass contain complex layers of electrodes, multiple organic films, terminals, and semi-conductors. The key materials for manufacturing TFT and CF glasses are the rare earth indium and valuable glass substrates with high heat resistance and high mechanical strength. The indium is employed in the form of ITO (indium tin oxide) [1] and is used for transparent electrodes due to its low resistance and patterning capability. However, indium is limited in production and its market price has increased [2]. Furthermore, an increasing quantity of LCD wastes is expected in a few years, thus it is essential to
∗ Corresponding author at: Department of Chemical and Environmental Engineering, Faculty of Engineering, Universiti Putra Malaysia, 43400 UPM Serdang, Selangor, Malaysia. E-mail address: [email protected] (H. Yoshida). http://dx.doi.org/10.1016/j.supflu.2015.05.016 0896-8446/© 2015 Elsevier B.V. All rights reserved.
recover precious resources such as indium from LCD panel wastes [3]. Efforts to recover valuable materials, especially indium from LCD wastes and ITO targets, have been investigated by ionexchange [4] and acid leaching [5,6]. Solvent extraction [7,8], hydrometallurgical [9] and chlorination [10,11] methods have also been reported to recover ITO from etching solutions or target. However, due to complications in the above processes, a simpler treatment method by means of sub-critical water (hereafter called sub-CW) technology was initiated by our group [12]. This technique is based on the use of water at temperatures between 100 ◦ C and 374 ◦ C, and at pressures high enough to maintain the liquid state. The magnitude of the ionic product of water increases three orders at around 250 ◦ C compared to room temperature [13]. This property is advantageous to facilitate hydrolysis and the decomposition of organic compounds. In addition, the dielectric constant of water decreases, thereby lowering its polarity. Thus, there are many reports on sub-CW for waste treatment and utilization [14–18]. In our previous paper [12] using sub-critical water at 360 ◦ C and a reaction time of 5 min, 83% indium recovery from CF glass was
H. Yoshida et al. / J. of Supercritical Fluids 104 (2015) 40–48
41
reported. Transparent glass was also recovered at the same time from CF glass. However, in the TFT glass, only about 10% of the indium was recovered. This low yield was due to the insufficient decomposition of organic matter from the TFT glass. The removal of the organic multilayer and the hydrolyzed layer required more aggressive decomposition reaction conditions than those employed in the previous study. The objectives of the present work are to obtain a higher indium recovery using sub-CW treatment and to reduce the reaction temperature. Suyama et al. found that basic materials, like alkylamines and organic compounds like long alkyl chained alcohol, used as additives in sub-CW are efficient to decompose polyesters and polystyrene [19–21]. In the present work, the possibility of using several basic inorganic hydroxides Na2 CO3 , ammonia and amine compounds in sub-CW is tested to increase the yield of indium and reduce the reaction temperature. At the same time, the recovery of clean glass substrate from the remaining treated LCD is also investigated.
2. Materials and experimental procedure 2.1. Preparation of LCD glass sample The materials and experimental procedure carried out in the present study were the same as those reported in our previous work [12]. In brief, the LCD panel used in this work was a 40 inch TFT type LCD panel, composed of TFT glass and CF glass. The polarizing films were removed from the panel. Then the panel was separated into CF glass and TFT array using a glass cutter (Toshin-Riko). Liquid crystals on both glass surfaces were washed away using acetone. Thereafter, both glasses were broken into about 5–10 mm sizes, small enough to fit and be in motion in the reactor. From here onwards, CF glass and TFT glass denote glass specimens without the liquid crystal layer and polarizing films.
2.2. Sub-CW water treatment The reaction field tested in this work was ultrapure water (MilliQ < 0.05 S cm−1 ) and aqueous solutions of NaOH, KOH, Na2 CO3 , Ca(OH)2 , ammonia and diethylamine (DEA), respectively. All reagents were analytical grade. Each aqueous solution was made using MilliQ water. The batch reactors used involved a stainless steel pipe (SUS316, 16 mm id, 150 mm length, 30 × 10−6 m3 ) and stainless steel tube end-locks (Swagelok). First, one side of the pipe was tightly sealed using the end-lock. After pouring 20 cm3 of each solution together with 6 g of either CF glass or TFT glass into the reactor, the air gap was replaced with argon gas. Thereafter, the reactor was sealed with the Swagelok cap and was then weighed. It was then immersed quickly in a preheated molten salt bath (Celsius, Thomas Kagaku Co. Ltd., Tokyo) containing a mixture of potassium nitrate and sodium nitrate with a stable temperature, and shaken. After a period of time, the reaction tube was taken out of the salt bath and immediately quenched in a tap water bath at room temperature. The temperature inside the reactor instantly increased until it equals the salt bath temperature [22]. The reaction time was defined as the period of time from immersing the reactor in the salt bath until transferring it into the tap water bath. The contents were recovered from the reactor and their physical changes were observed. The solid residues were separated from the aqueous phase by filtration through a 1.0 m pore-sized membrane filter (Advantec, cellulose acetate). The reactor tube and cap were rinsed with MilliQ water and the water was filtered together with the aqueous phase.
Fig. 1. Effect of NaOH, KOH, Na2 CO3 , DEA, Ca(OH)2 and NH3 in sub-CW on indium recovery. All additive concentrations were 0.1 N. Reaction temperatures and time was 260 ◦ C and 5 min, respectively.
2.3. Measurement and analysis After sub-CW treatment, the metal content in each medium was determined as follows: the indium on the glass and on the membrane filter was extracted using a known amount of 7% HCl aq. solution, respectively, according to Eq. (1): In2 O3 + 6HCl → 2InCl3 + 3H2 O(1) Then they were sonicated for 1 h at 50 ◦ C. The indium concentration in the HCl aq. solution was measured using ICP (Shimadzu, ICPE-9000) and the amounts of indium in the fresh glass (W0 ) and in the remaining glass substrate (Wg ), filter (Wf ) and liquid-phase (Wl ) were determined. The material balance of indium before and after the reaction was confirmed. The recovery of indium (in ) was calculated using Eq. (2): in =
Wf × 100 W0
(2)
Subsequent to the indium analysis, Sn, Si, and Fe concentrations were also determined from each sample by ICP. The total organic carbon content in the liquid-phase was measured using a TOC analyzer (Shimadzu, TOC-Vcph/cpn). The metallurgical microstructure images of the glass surface were obtained using an optical fluorescence microscope (BX51 Olympus Corp.). 3. Results and discussion 3.1. Effect of basic materials in sub-CW Our previous paper [12] showed that the maximum recovery of indium from TFT glass by sub-CW was only 10%. To increase the recovery, the effect of basic materials in sub-CW on indium recovery was tested. Fig. 1 shows the effect of the basic material in sub-CW on the recovery of indium from TFT glass at 260 ◦ C for 5 min. Without basic materials (water only), only 5% of the indium was recovered. However, when sub-CW contained 0.1 N NaOH, KOH, Na2 CO3 and Ca(OH)2 , respectively, the indium recovery increased to 90%, 85%, 60%, and 25%, respectively. The effects of 0.1 N NH3 and DEA were also investigated and resulted in an indium recovery of 2% and 8%, respectively. Sub-CW treatment with the presence of alkali increased the indium recovery. Alkali which contained OH− demonstrated a higher indium recovery than without
42
H. Yoshida et al. / J. of Supercritical Fluids 104 (2015) 40–48
the hydroxide ion. Suyama et al. [19–21] reported that amines were effective to improve hydrolysis. In this study, KOH and NaOH showed superior indium recovery compared to other additives because they are strong bases due to their smaller pKb (0.5 and 0.2, respectively). In addition, Ca(OH)2 , DEA and NH3 in which their pKb are 2.4, 3.2, and 4.7, respectively, are weaker base and thus resulted in less indium recovery. Consequently, the result showed a good relation between the pKb and the indium recovery. Since the addition of NaOH produced the best indium recovery and the price of NaOH is cheaper than KOH, only NaOH was used in further studies in this work. 3.2. Sub-CW treatment of TFT glass Fig. S1 (in Supplemental information) demonstrates photographs of the TFT glass after treatment of sub-CW with 0.1 N NaOH for 5 min at reaction temperatures between 120 and 360 ◦ C. Below 160 ◦ C, no change of color was observed in the aqueous solution, but at 180 ◦ C it became slightly brown. At 200 ◦ C, brownish floating materials, which were possibly the organic film, were observed in large amounts. At around 240 ◦ C, the color of the solution transformed from light brown to intensely brown. This is because the organic film on the surface of the TFT glass was hydrolyzed due to the sub-CW reaction and then changed to some substances that were water soluble. The aqueous solution was then filtered through the membrane filter. Fig. S2 (in Supplemental information) shows the photographs of the filters. At 200 ◦ C and 220 ◦ C, the brownish water-insoluble materials were observed on the filter. They were the organic multilayer films wrapped together with ITO exfoliated from TFT glass. This phenomenon suggested that the leaching of indium did not occur because the solid film layers did not dissolve in the aqueous solution. The concentration of indium in the aqueous phase is given below. However, above 240 ◦ C, the organic film became smaller, and started to shrink above 260 ◦ C due to the decomposition of the organic material. 3.3. Sub-CW treatment of CF glass Fig. S3 (in Supplemental information) shows photographs of the CF glass after treatment of sub-CW with 0.1 N NaOH between 100 and 360 ◦ C for 5 min. At 100 ◦ C, the aqueous solution was clear and no physical change was observed. However at 120 ◦ C, the color of the solution was yellow, before becoming extremely yellow at 140 ◦ C. Furthermore, dye pigments consisting of red, green, and blue were observed floating in large amounts in the solution. At temperatures above 160 ◦ C, the solution changed to dark blue before becoming dark brown at 320 and 360 ◦ C. After sub-CW treatment, the aqueous solution was filtered through the membrane filter. Photographs of the membrane filter surface after the filtration are shown in Fig. S4 (in Supplemental information). At 100 ◦ C, nothing was detected on the filter. The release of substantial amounts of water-insoluble green pigments were observed at 120 ◦ C, then red pigments at 140 ◦ C, and blue pigments at 160 ◦ C. The color became dark above 180 ◦ C. These phenomena showed that the exfoliation of the color pigments occurred in subsequent steps from green, red, and blue. As a consequence, these filtered water-insoluble CF materials were present due to the decomposition of the organic films on the CF glass surface caused by the hydrolysis reaction in sub-CW with 0.1 N NaOH. 3.4. Recovery of indium from TFT and CF glasses In order to evaluate the amount of indium recovery by subCW reaction, the amounts of Wg , Wf and Wl were determined and
Fig. 2. Effect of reaction temperature on presence of indium in each phase after treatment of sub-CW with 0.1 N NaOH. (a) TFT glass and (b) CF glass. Reaction time of 5 min.
expressed in milligrams of indium per kilogram of glass substrate (mg/kg-GL). Fig. 2(a) shows the results after treatment of TFT glass by sub-CW with 0.1 N NaOH. An excellent mass balance of indium quantity that was distributed in each phase was obtained. Indium was not detected at any temperatures in the liquid-phase. Treatment below 160 ◦ C showed no decrease of Wg , denoting that no indium was removed from TFT glass. However, above 160 ◦ C, Wg decreased sharply from 290 mg/kg-GL to 16 mg/kg-GL (200 ◦ C). This suggests that a large amount of indium was recovered from the TFT glass surface. Wg remained below 10 mg/kg-GL at 220–320 ◦ C but increased to 70 mg/kg-GL at 360 ◦ C. Meanwhile, Wf showed a trend opposite to Wg . Wf was zero below 160 ◦ C, but then increased sharply at 180–200 ◦ C. At 220 ◦ C, Wf was the highest with almost 290 mg/kg-GL. At 230–320 ◦ C, Wf was a little smaller than the highest value. However, Wf decreased when the reaction temperature was above 340 ◦ C, indicating that some indium remained on the TFT glass. Fig. 2(b) shows the results in CF glass. Indium was not detected in the liquid-phase. Even at 100 ◦ C, Wg reduced to 162 mg/kgGL. At 140 ◦ C, Wg significantly decreased to below 5 mg/kg-GL. It remained below 5 mg/kg-GL between 140 and 320 ◦ C. However at 360 ◦ C, Wg slightly increased to 40 mg/kg-GL. Wf showed the opposite trend to Wg . At 100 ◦ C, Wf was 170 mg/kg-GL. It increased to 350 mg/kg-GL at 140 ◦ C. At reaction temperatures between 140 ◦ C and 320 ◦ C, Wf remained high (between 340 and 360 mg/kg-GL) but slightly dropped to 300 mg/kg-GL at 360 ◦ C. Thus, treatment of sub-CW with 0.1 N NaOH between 140 and 320 ◦ C resulted in a complete exfoliation of CF materials, including indium, from the CF glass and they could be completely recovered by the membrane filter.
H. Yoshida et al. / J. of Supercritical Fluids 104 (2015) 40–48
43
3.5. Optic microscope images of TFT and CF glasses
Fig. 3. Effect of NaOH on the indium recovery after sub-CW reaction over (a) TFT and (b) CF glasses for 5 min at various temperatures. The solid and dotted lines indicate sub-CW reaction with 0.1 N NaOH and without NaOH [12], respectively.
The reason why Wg increased and Wf decreased in both TFT and CF glasses when the reaction temperature was higher than 340 ◦ C will be discussed in a later section. Fig. 3 illustrates the effect of reaction temperature on the indium recovery (in ) after treatment of sub-CW with and without 0.1 N NaOH. Fig. 3a shows the results in TFT glass. The dotted line shows the results without NaOH and in was only less than 10% [12]. However, when 0.1 N NaOH existed in sub-CW and the reaction temperature was higher than 160 ◦ C, in increased to 95% at 220 ◦ C. The in was around 93% between 220 and 320 ◦ C. This shows that the treatment of sub-CW with 0.1 N NaOH was far superior compared to that without NaOH. TFT substrates are usually coated with a thin (∼80–100 nm) alignment layer, insulating the optomer layer and protecting the passivation layers (see Fig. S1 in [12]). The optomer layer is an organic mixture containing mainly diethylene glycol methyl ethyl ether (MEC) and acrylic resin. Meanwhile, the passivation layer covering the drain and source terminals are made of SiNx . This suggested that the decomposition of the optomer and passivation layers on the TFT glass by sub-CW was enhanced drastically by the presence of NaOH. In the CF glass in Fig. 3(b), the effect of NaOH was also clear. When 0.1 N NaOH existed in sub-CW, in was 45% even at 100 ◦ C. This means that the hydrolysis reaction was active and had already taken place even at the water boiling point. in increased to 99% at 160 ◦ C. Between 160 and 320 ◦ C, in was somewhere between 90 and 99%. The dotted line shows the result without NaOH [12], where it was much smaller than that with 0.1 N NaOH. CF glass consists of 2–3 m thin color pigments, an overcoat layer and a 1 m thin black matrix [23,24]. These organic substances may be more susceptible to attacks from OH− in sub-CW. Thus exfoliation of these multiple organic layers occurred even at 100 ◦ C, which was lower than that in TFT glass.
Fig. 4 shows TFT glass surfaces observed using an optical microscope before and after treatment of sub-CW with 0.1 N NaOH. No difference was observed for (a) before reaction and (b) treated at 120 ◦ C. In (c) treated at 160 ◦ C, it became slightly distorted and showed early signs of exfoliations of the organic multilayer film. At 200 ◦ C (d), the glass surface became partially visible and this showed that the organic multilayer film was stripping away. The materials on the membrane filter were therefore attributed to the presence of the exfoliated organic multilayer films (Fig. S2). At 220 ◦ C (e), a clear transistor surface was observed with the gate and source bus-line terminals still remaining in an intact condition, and the whole organic multilayer film was exfoliated as evidenced from the filtered material (Fig. S2). This led to the extremely high indium recovery at 220–320 ◦ C as observed in Figs. 2 and 3. At 360 ◦ C (f), the source and gate bus-lines were slightly detached and grayish matter appeared on the surface as if a layer was formed. Fig. 5 shows optical microscope images of the CF glass before reaction (a) and after treatment of sub-CW with 0.1 N NaOH. At 100 ◦ C (b), the pigment colors became paler compared to (a). At 120 ◦ C (c), 140 ◦ C (d) and 160 ◦ C (e), exfoliations of green pigment followed by red and blue pigments were observed. This agrees with the results in Fig. S4, where the filtered material was green, combined red–green and combined blue–red–green at 120 ◦ C, 140 ◦ C, and 160 ◦ C, respectively. The black matrix was slightly visible at 160 ◦ C (e). At 280 ◦ C (f), the glass was transparent, without the color pigments and black matrix. This indicates that the alignment film, overcoat layer, ITO and color filters were totally exfoliated from the CF glass. This was clear in Fig. S4, where the filtered material was completely dark, with a tremendously high indium recovery between 180 and 280 ◦ C. 3.6. Mechanism of exfoliation of multi-layered organic membranes from LCD The exfoliation mechanisms in sub-CW with 0.1 N NaOH are further discussed in Fig. 6. In the TFT glass (Fig. 6(a)), the multi-layered organic membranes that consist of alignment, optomer and passivation layers were not removed below 160 ◦ C, but to some extent were removed at 200 ◦ C. Since ITO was located below the alignment film, the ITO exfoliated from TFT glass during the sub-CW treatment was in a state which was sandwiched between the organic films. Thus, the indium recovery at 200 ◦ C is ascribed to the partial exfoliation of the organic multilayer film. However at 220 ◦ C, a complete exfoliation of the layers occurred, thus the whole of the layers including ITO was removed because of the strong hydrolysis of sub-CW with 0.1 N NaOH. This strong hydrolysis is explained by the maximum ionic product of water at around 250 ◦ C as shown in Fig. S5 (in Supplemental information). Fig. 6(b) illustrates the mechanism of exfoliation of the multiple layers from CF glass. The pale color of the microscopic figure could be due to the partial removal of the alignment layer and ITO at 100 ◦ C, where about 50% of them may have been removed. Thus, indium recovery was about 50%. At 120, 140, and 160 ◦ C, the color pigments of red, green, and blue, respectively, were removed in turn from the surface but the black matrix still remained on it. At 280 ◦ C, all of the alignment and ITO layers, the color pigments and the black matrix were exfoliated from the glass. This was confirmed by the transparent glass photo from the microscopic image in Fig. 5(f) and the color of the filtered material in Fig. S4. Fig. 2 showed that Wg at 360 ◦ C was larger than those at lower temperature. This phenomenon is related to the fact that the hydrolysis activity shows a maximum at around 250 ◦ C and gets weaker when the reaction temperature is higher than 300 ◦ C (Fig.
44
H. Yoshida et al. / J. of Supercritical Fluids 104 (2015) 40–48
Fig. 4. Microscope images of (a) untreated TFT glass, and the TFT glass treated by sub-CW with 0.1 N NaOH for 5 min at (b) 120 ◦ C, (c) 160 ◦ C, (d) 200 ◦ C, (e) 220 ◦ C, and (f) 360 ◦ C, respectively.
Fig. 5. Microscope images of the (a) untreated CF glass, and CF glass treated by sub-CW with 0.1N NaOH for 5 min at (b) 100 ◦ C, (c) 120 ◦ C, (d) 140 ◦ C, (e) 160 ◦ C, and (f) 280 ◦ C, respectively.
S5). As at 360 ◦ C, the hydrolytic strength becomes weaker than that at 250 ◦ C, the mechanisms of exfoliation in TFT and CF glasses are explained as Fig. 6(A) and (B) at 360, respectively. 3.7. TOC and Si concentrations of the liquid-phase TOC was measured to make clear how much water-soluble organic material in the liquid-phase was produced by sub-CW with 0.1 N NaOH. Fig. 7 shows the effect of the reaction temperature
on TOC in the liquid-phase. TOC concentrations from both TFT and CF glasses clearly are much higher than those without NaOH [12]. They escalated with increasing temperature. In TFT glass, the TOC concentration was zero at T < 180 ◦ C. TOC increased sharply to 1.34 mg/g-GL in 180 < T <260 ◦ C. At 260 < T < 360 ◦ C, TOC was about 1.34 mg/g-GL. On the other hand, in the CF glass, TOC concentration was 0.2 mg/g-GL even at 100 ◦ C and increased sharply to 1.13 mg/gGL at 180 ◦ C, then gradually increased and was close to that of TFT glass at T > 260 ◦ C.
H. Yoshida et al. / J. of Supercritical Fluids 104 (2015) 40–48
45
Fig. 6. Mechanism of exfoliation of organic multiple layer from TFT and CF glasses during treatment of sub-CW with 0.1 N NaOH at various temperatures (Reaction time: 5 min).
The TOC concentration in CF is higher than that in TFT in T < 250 ◦ C. This is evidence that the organic layers in CF and TFT glasses are different, and thus would require different reaction temperatures for exfoliation from the glasses. To ascertain the exfoliation of the passivation layer that consisted of SiNx , the silica concentration in the liquid-phase was determined by ICP by a similar extraction method to when determining the indium concentration. Fig. 8 shows Si concentration in each phase after TFT and CF glasses were treated by sub-CW with 0.1 N NaOH. In the TFT glass, Si concentration was less than 500 mg/kg-GL in all phases when treated below 190 ◦ C. However,
Si concentration in the liquid-phase increased sharply to about 3800 mg/kg-GL between 280 and 320 ◦ C. This suggested that SiNx from the passivation layer has dissolved into the liquid-phase as a result of the exfoliation of the organic multi-layers with ITO. On the other hand, the increase of Si concentration was not observed in the CF glass. This is reasonable since CF glass does not contain any passivation layer. Above 300 ◦ C, the silica concentrations on the glass increased in both TFT and CF glasses. This is probably due to the corrosion of the substrate glass in NaOH environment at a very high reaction temperature causing the formation of sodium silicate.
46
H. Yoshida et al. / J. of Supercritical Fluids 104 (2015) 40–48
Fig. 7. Effect of temperature on total organic carbon (TOC) concentrations of liquid phase after treatment of TFT and CF glasses by sub-CW with 0.1 N NaOH (reaction time: 5 min). Dotted line indicates the sub-CW treatment without NaOH [12].
Fig. 9. NaOH concentration of the resultant solution as a function of reaction temperature. Reaction time: 5 min.
Fig. 8. Presence of silica in each phase during sub-CW reaction with 0.1 N NaOH and reaction time of 5 min as a function of reaction temperatures. (a) TFT glass and (b) CF glass.
Fig. 9 illustrates the NaOH concentration of the resultant solution that was measured by titration. The NaOH concentration from TFT glass was 90 mM from 120 ◦ C until about 270 ◦ C before it dropped sharply to 20 mM (360 ◦ C). The NaOH concentration from CF glass dropped steadily from 90 mM (100 ◦ C) to 85 mM (270 ◦ C) before it sharply dropped to 20 mM (360 ◦ C). This supports the evidence that SiNx from the passivation layer has dissolved into the liquid-phase as a result of the exfoliation of the organic multilayers. This is followed by corrosion of the substrate glass in NaOH environment above 270 ◦ C causing the formation of sodium silicate. 3.8. Effect of NaOH concentration and reaction time. Fig. 10 shows the effect of the concentration of NaOH on indium recovery. The reaction time was 5 min while the reaction temperatures were taken from the optimum reaction temperatures for TFT and CF glasses which were 220 and 160 ◦ C, respectively. No increase in in was observed between 0.003 and 0.01 N in both TFT and CF glasses. When the concentration was above 0.01 N, in increased sharply and was above 95% at 0.1 N. in of both glasses remained almost constant above 0.1 N. This clearly shows that 0.1 N NaOH is the optimum for indium recovery from both TFT and CF glasses.
Fig. 10. Effect of NaOH concentration on indium recovery over (a) TFT glass at reaction temperature of 220 ◦ C and (b) CF glass at 160 ◦ C (Reaction time: 5 min).
The optimum reaction time was investigated in 0.1 N NaOH for TFT and CF glasses at 220 ◦ C and 160 ◦ C, respectively. The results are given in Fig. 11. in in TFT and CF glasses increased sharply to about 80% in only 2 min. At 5 min, in reached 95% for TFT glass and 99% for CF glass. No further increase was observed for reaction times longer than 5 min. This indicates that a reaction time of 5 min is sufficient for both TFT and CF glasses.
H. Yoshida et al. / J. of Supercritical Fluids 104 (2015) 40–48
47
was recovered just by filtration, which is a much simpler recovery process than acid dissolution, chlorination and ion-exchange methods. The amount of indium recovered from CF glass and TFT glass was dependent on the reaction temperature. The optimum temperatures for TFT and CF glasses were 220 ◦ C and 160 ◦ C, respectively. The indium recoveries from TFT and CF glasses were 95% and 99%, respectively. The minimum requirements for indium recovery were 0.1 N for NaOH concentration and 5 min of reaction time. The dissolution of the Fe ion from the reactor wall was small, demonstrating the vessel wall can withstand sub-critical conditions with 0.1 N NaOH. With the low cost and low corrosion effect of NaOH, the treatment of sub-CW with NaOH has high potential to be applied in the recovery of indium from LCD panel wastes. Appendix A. Supplementary data Supplementary data associated with this article can be found, in the online version, at http://dx.doi.org/10.1016/j.supflu.2015.05. 016 References
Fig. 11. Effect of reaction time on indium recovery over (a) TFT glass and (b) CF glass at reaction temperature of 220 ◦ C and 160 ◦ C, respectively.
3.9. Corrosion of the reactor wall Concerning its commercial application, the sub-CW apparatus should be used for a long period. The corrosion of the surface of the reactor in sub-CW in the presence of 0.1 N NaOH was tested. Fig. S6 shows the dissolution of the Fe ion from the internal surface of the batch reactor (SUS316) at 260 ◦ C. During the first 50 min, the Fe ion increased sharply from 0 to 0.28 mg Fe/20 mL NaOH. However beyond 50 min, the Fe ion increased slightly to 0.35 mg Fe/20 mL NaOH. This showed that the Fe dissolution rate of the vessel wall was tremendously slow. Assuming a homogeneous dissolution of the wall, it is predicted that the wall thickness lessens by a mere 2 m a year. Thus, due to this small effect, it is considered that vessel walls can withstand continuous and extensive operation when treated by sub-CW in the presence of 0.1 N NaOH. 4. Conclusions Indium in the form of ITO was recovered from TFT and CF glasses by treatment of sub-CW in additives. In the presence of NaOH, KOH, Na2 CO3 , Ca(OH)3 , and NH3 , respectively, NaOH gave the highest recovery of indium. Sub-CW treatment in 0.1 N NaOH resulted in a near complete recovery of indium from TFT and CF glasses in just a short period of 5 min. This result was a remarkable progress from our previous study using only water [12]. ITO was recovered from the glass together with the organic multi-layers that did not dissolve in the liquid-phase. This is very beneficial because indium
[1] A.M. Alfantazi, R.R. Moskalyk, Processing of indium: a review, Min. Eng. 16 (2003) 687–694. [2] G. Phipps, C. Mikolajczak, T. Guckes, Indium and Gallium: long-term supply, Renew. Energy Focus 9 (2008) 56–59. [3] S. Stalhofer, M. Spitzbart, K. Maurer, Recycling of LCD screens in Europe – State of the art and challenges, in: J. Hesselbach, C. Herrmann (Eds.), Globalized Solutions for Sustainability in Manufacturing, Springer-Verlag Berlin Heidelberg, 2011, pp. 454–458. [4] M. Tsujiguchi, Design for innovative value towards sustainable society, in: M. Matsumoto, Y. Umeda, K. Masui, S. Fukushige (Eds.), Proceedings of EcoDesign 2011: 7th International Symposium on Environmentally Conscious Design and Inverse Manufacturing, Springer, Dordrecht, 2012, pp. 743–746. [5] J. Li, S. Gao, H. Duan, L. Liu, Recovery of valuable materials from waste liquid crystal display panel, Waste Manage. 29 (2009) 2033–2039. [6] Y. Li, Z. Liu, Q. Li, Z. Liu, L. Zeng, Recovery of indium from used indium–tin oxide (ITO) targets, Hydrometallurgy 105 (2011) 207–212. [7] M.C.B. Fortes, A.H. Martins, J.S. Benedetto, Indium recovery from acidic aqueous solutions by solvent extraction with D2EHPA: a statistical approach to the experimental design, Sharp Tech. J. 92 (2005) 17–22. [8] T. Honma, T. Muratani, Material collection from liquid crystal display wasted panels, Sharp Tech. J. 92 (2005) 17–22. [9] S.J. Hsieh, C.C. Chen, W.C. Say, Process for recovery of indium from ITO scraps and metallurgic microstructures, Mat. Sci. Eng. B 158 (2009) 82–87. [10] K.S. Park, W. Sato, G. Grause, T. Kameda, T. Yoshioka, Recovery of indium from In2 O3 and liquid crystal display powder via a chloride volatilization process using polyvinyl chloride, Thermochim. Acta 493 (2009) 105–108. [11] K. Takahashi, A. Sasaki, G. Dodbiba, J. Sadaki, N. Sato, T. Fujita, Recovering Indium from the liquid crystal display of discarded cellular phones by means of chloride-induced vaporization at relatively low temperature, Metall. Mater. Trans. A 40 (2009) 891–900. [12] H. Yoshida, S. Izhar, E. Nishio, Y. Utsumi, N. Kakimori, S.A. Feridoun, Recovery of indium from TFT and CF glasses in LCD panel wastes using sub-critical water, Solar Energy Mater. Solar Cells 125 (2014) 14–19. [13] W. Wagner, H.-J. Kretzschmar, International Steam Tables: Properties of Water and Steam, 2nd ed., Springer-Verlag Berlin Heidelberg, 2008, pp. 322. [14] H. Yoshida, H. Tokumoto, K. Ishii, R. Ishii, Efficient, high-speed methane fermentation for sewage sludge using subcritical water hydrolysis as pretreatment, Bioresour. Technol. 100 (2009) 2933–2939. [15] O. Pourali, F.S. Asghari, H. Yoshida, Production of phenolic compounds from rice bran biomass under subcritical water conditions, Chem. Eng. J. 160 (2010) 259–266. [16] O. Tavakoli, H. Yoshida, Application of sub-critical water technology for recovery of heavy metal ions from the wastes of Japanese scallop Patinopecten Yessoensis, Ind. Eng. Chem. Res. 45 (2006) 5675–5680. [17] W. Abdelmoez, H. Yoshida, Kinetics and mechanism of the synthesis of a novel protein-based plastic using subcritical water, Biotechnol. Prog. 24 (2008) 466–475. [18] W. Abdelmoez, H. Yoshida, Mechanical and thermal properties of a novel protein-based plastic synthesized using subcritical water technology, Macromolecules 40 (2007) 9371–9377. [19] K. Suyama, M. Kubota, M. Shirai, H. Yoshida, Effect of alcohols on the degradation of crosslinked unsaturated polyester in sub-critical water, Polym. Degrad. Stab. 91 (2006) 983–986. [20] K. Suyama, M. Kubota, M. Shirai, H. Yoshida, Degradation of crosslinked unsaturated polyesters in sub-critical water, Polym. Degrad. Stab. 92 (2007) 317–322.
48
H. Yoshida et al. / J. of Supercritical Fluids 104 (2015) 40–48
[21] K. Suyama, M. Kubota, M. Shirai, H. Yoshida, Chemical recycling of networked polystyrene derivatives using subcritical water in the presence of an aminoalcohol, Polym. Degrad. Stab. 95 (2010) 1588–1592. [22] W. Abdelmoez, H. Yoshida, Simulation of fast reactions in batch reactors under sub-critical water conditions, AICHe J. 52 (2006) 3600–3611.
[23] T.M. Lee, Y.J. Choi, S.Y. Nam, C.W. You, D.Y. Na, H.C. Choi, D.Y. Shin, K.Y. Kim, K.I. Jung, Color filter patterned by screen printing, Thin Solid Films 516 (2008) 7875–7880. [24] P.M. Harrison, J. Wendland, M. Henry, Innovative laser patterning of black,atrix for LCD manufacture, Soc. Infor. Display (SID) Symp. Digest 39 (736) (2008).