Evaluation of Anode Water Electrolyzed with Anion Exchange Membrane for Cleaning EUV Semiconductor

Electrically nonconducting UPW was electrolyzed without electrolyte through anion exchange membrane for evaluating applicability to EUV semiconductor cleaning. Produced anode water held positive ORP up to 900 mV, which is very oxidative. ORP, pH, and conductivity measurements showed delicately complementary each other for understanding anode water. Correlation of concurrent ORP decrease and conductivity increase in ultra-pure anode water domain was observed first time. The oxidative OH° was formed as the major species in anode water, causing positive ORP during ORP measurement. H+ and OH− ions, and OH° radical coexisted in anode water at amphoteric nonequilibrium, while pH was less than 6. It was concluded that OH°, as a strong oxidant, transformed itself to OH− by ORP measurement. OH° radical would oxidize selectively and then remove nano-contaminants. Anode water is considered to fulfill the requirement of EUV semiconductor cleaning where no oxygen species should be required because of likely oxide layer formation during cleaning, and it will even remove the native oxide developed unintentionally before cleaning.

The wet cleaning has been shifted to the non-chemical electrolyzed water cleaning in EUV(extreme ultraviolet) semiconductor technology.There appear very various material surfaces to be cleaned at the same time in front and backend of EUV processes such as metals, compounds, and dielectrics of low and high k.Even dielectrics diversify to various chemical forms of hydrides, hydrates, oxides, nitrides, and so on.Hence, this wet cleaning requirement is now more stringent before drying nano-scale devices.Cathode water has already been successfully applied to EUV semiconductor cleaning of non-chemical requirement. 1,2Even though anode water is also an excellent alternative for such as cleaning or rinsing, it has not been examined comprehensively.UPW(ultrapure water) and electrolyzed water are highly insulating liquids, and they are not fully understood for explaining their quality, properties, behavior, and cleaning reactions, by monitoring basically ORP(oxidation reduction potential), pH, and conductivity measurements.Cleaning is practically dependent upon only these three parameters.Even though these three measurements contain very fundamental information, they have not been still correlated comprehensively together to extract out the meaningful comprehensions of UPW or electrolyzed water cleaning so far.
As well known, ORP measures voltages developed by galvanic reaction between two electrodes of Ag/AgCl reference and Pt in measured solution.As soon as Pt electrode of ORP meter immerges in solution for measurement, the electromotive force is developed and electron flow starts in one way or another between Ag/AgCl electrode in reference solution and Pt electrode depending upon chemical species in measured solution.Ag may be oxidized or reduced, while chemical species on Pt electrode surface are reversely reduced or oxidized, causing the electromotive force through the close circuit among Pt electrode, Ag electrode, and the measured solution.Therefore, ORP meter measures the voltage and its directionality developed by ongoing reaction between chemical species in solution and Ag/AgCl electrode, and not developed voltage by electrostatically and spatially rearranged chemicals like hydrogen ions in pH meter.Very importantly, as a result, ORP measurement itself alters chemical state of the solution as measuring time goes on, while pH measurement does not.Meanwhile, pH meter measures literally the voltage developed on the sensor electrode by only hydrogen ions(proton) in solution because it becomes physically an electronic device in silicic acid glass doped with Li just like MOS(metal oxide semiconductor). 3,4Hence, pH meter is designed inherently to monitor only hydrogen ion concentration changes, and unfortunately not to visualize other chemical species in solution.6][7][8] These different mechanisms of two measurements provide the delicate key for understanding UPW where no other chemical species exists except H + and OH − ions, and give very fundamental and complementary parameters together for evaluation of non-chemical solutions such as electrolyzed water.Conductivity, the third measurement, is a result of flow of all free charged species passing across between two measuring electrodes in solution.Therefore, conductivity cannot differentiate or separate species as well as their directionalities like ORP measurement, and is considered as a sum of all ionic species in UPW or electrolyzed water.Hence conductivity can be a candidate as a cross-checking and complementary tool because it changes by itself due to charged species newly formed by ORP measurement reaction.However, there has been no such correlation for UPW or electrolyzed water of highly electrical insulator. 9,10UPW can be produced through two steps that membrane filtration pushes conductivity down around 1 μS cm −1 , then followed by electrode-ionization ultimately achieving 0.05 μS cm −1 or less, which is equivalent to resistivity of 18.2 Mohm. 11Since the conductivity measurement can detect the level of 0.05 μS cm −1 or higher, it is useful for materializing delicate chemical variations in UPW, and electrolyzed water as well. 12rom the above discussions, it can be summarized that pH, OPR, and conductivity measurements of UPW together at the same time are mutually complementary, which can provide the comprehensive understandings of very extremely noble aqueous liquid.UPW can be said to be noble since it is so pure and insulating that no other chemical element is involved other than hydrogen and oxygen. 13,14easurement of ORP, pH, and conductivity was introduced in this study and compared each other since it was sufficiently enough to understand the anode water produced with AEM(anion exchange membrane) located between two electrolyzing electrodes.It is also stressed that this work will propose a promising solution for EUV z E-mail: bschoi@ewtechnology.co.krECS Advances, 2023 2 040510 cleaning process where non-chemical UPW including anode water should be inevitable, as well as surface treatment applications of nano-surfaces of inorganics and organics in future.

Experimental
Self-assembled equipment of Neo-radical5 for the electrolysis experiment was optimized and produced by EW technology, as shown in Fig. 1.The electrolysis cell was designed in such a way that the AEM was located between cathode and anode.The incoming UPW was supplied with ratio of about 4:1 for anode and cathode sides, respectively.UPW flow rates and electrolysis currents were in the range of 100-500 ml min −1 and 2-5 amperes(A), respectively.There were three measurement equipments which were TOA DKK HM-31P for ORP and pH, and Greisinger GMH 3431 for conductivity.
As the produced anode water was filled in 300 ml glass bottles of several groups, depending on various measurement procedures.Measurements(M) were carried out as shown in Fig. 2, after assuring that any likely electrostatic charge in anode water was eliminated. 2 For the first group for measuring ORP and conductivity, the starting step was to dip in ORP electrode, then to collect ORP data, and to take ORP electrode out from the glass bottle.Then, ORP electrode was dipped in again after 5 min elapse(dot line) followed by immediate ORP measurement, then held for 5 min(solid line), and taken out after ORP measurement.This sequence continued repeatedly for 180 min (phase I).For the next group, ORP data were collected in every 10 min repeatedly for 180 min while keeping ORP electrode dipped in glass bottles all the way(phase II).The reason to set up these two procedures was to observe any likely ORP difference due to measurement time between two groups since ORP data should take place theoretically due to total time differences of ORP measurement reactions.In parallel, conductivity measurements were carried out just after every ORP measurement to observe the results of ORP measurement reactions, since conductivity measurements could also reveal the influence of ORP measurement time differences, if any.In the third and fourth groups, same procedure of phase I was repeated for ORP measurement followed by pH measurement instead of conductivity measurement in order to observe influence of ORP measurements on pH.Finally, there were some complementary measurements specifically to confirm pH variations in some condition of no reaction of ORP measurement, by simply dipping and measuring only pH meter every 10 min for 180 min repeatedly without disturbance by dipping ORP meter at all.

Results and Discussion
Electrostatic charges of produced electrolyzed waters were removed before further experiments.Figure 3 showed the typical discharging effects of waters electrolyzed with CEM(cation exchange membrane).This electrostatic charge was caused by the highly electrical-insulating property of pure and electrolyzed water.After discharging, ORP approached to −600 and 500 mV ranges for cathode and anode waters(red data) electrolyzed with CEM, respectively.Anode water with AEM, though, improved ORP up to 900 mV as shown in Fig. 4.
Figure 4 showed ORP and pH variations of anode water with AEM for various electrolyzing currents and flow rates of UPW.ORP variations in Fig. 4a were only dependent on electrolysis currents, with little dependence on flow rates.ORP increased linearly as currents increased.It was also clear that increase of electrolysis currents resulted in increase of oxidative chemical formation to be reduced during the following ORP measurement, and that AEM even enhanced ORP up to 900 mV.More importantly, ORP was positive, which meant that the very strong oxidative chemical was produced in anode water during electrolysis.The strong oxidative chemical then was oxidizing the reference electrode and being transformed itself to the reduced form by gaining an electron at the same time during ORP measurement procedures.Actually, Ag of Ag/AgCl reference electrode was being oxidized by losing the corresponding electron.Hence, the negatively-charged reduced form should change the conductivity of the solution because of negative ion increase in solution due to continuously receiving electrons.
Another feature was that pH values seemed to decrease somehow as electrolysis currents increased, converging to a certain level, while it had nothing to do with flow rates, as shown in Fig. 4b.There could be a likely explanation of slight pH decrease and convergence, which was the increase of H + ions due to electrolysis current increase.The main reaction at anode electrode was the generation reactions of oxygen or ozone gases while conversing the produced OH − ion to water.This reaction did not mean that any left OH − ion concentrations after gas conversion decreased depending on the increase of electrolytic currents.Therefore, it was sure that there was some significant concentration increase of H + ions by electrolytic current increase as expected.Figure 5 showed property changes of electrolyzed waters as ORP measurement time increased.Most of all, anode water was degraded by ORP measurement relatively faster than cathode water as shown in Figs.5a and 5b.It took several days for cathode water to degrade 1 while only hours for anode water.It meant that anode water showed very strong reactivity, since it might contain the strong oxidant developed during electrolysis while cathode water did not.Actually, oxidant in cathode water was H + ions, while oxidant in anode water seemed to be quite strong.This is explained by the fact that reduction potential of H + ions is even weaker than that of Ag + ions in reference cell.Secondly, it showed very interesting ORP and conductivity correlations in anode water electrolyzed with AEM in case of ORP measurement time differences for phase I and II of the first and second groups as shown in Fig. 5b.It was clear that ORP measurement reaction for phase II deteriorated anode water much faster than that of phase I while conductivity of phase II increased faster than that of phase I, as time increased.This phenomenon explained very well about effect of ORP measurement reaction that the strong oxidant was being degraded and negatively-charged ions were being formed.This phenomenon was observed in anode water for the first time in this work.Thirdly, conductivity of anode water at initial stage of ORP measurement was 0.05 μS cm −1 or less, and did not show any significant increase at the initial 20 min of measurements even with electrolysis current increase.At this stage, H + ion concentration might increase a little due to AEM, but give no meaningful contribution to conductivity increase even though pH decreased somehow, as in Figs. 6 and 7. Therefore, it was clear that anode water maintained the insulating behavior even though nonequilibrium concentrations of H + ions and OH − ions coexisted amphoteric.
It was shown in Fig. 6 that ORP and pH measurements revealed clearly their independent relationship for phase I.As ORP measurements proceeded, ORP and pH values decreased.For pH, its variation seemed to be vigorous more or less at the initial measurement, but to be stabilized at about 6 as in Fig. 7.As ORP measurements proceeded, anode water deteriorated and OH − ion concentration seemed to increase because of conductivity increase.Total H + ion concentration also increased and pH decrease due to AEM as well as electrolysis currents, as shown in Fig. 4.Even though equilibrium constant of these two ions was to follow Le Chatelier's principle, they did not re-store equilibrium quickly enough due to the electrical insulating behavior of anode water.Therefore, the total pH decrease of non-equilibrium should be caused by accumulation of H + ions due to both AEM and ORP measurements.At the same time, some ion species such as OH − ion seemed to be formed and increase fast enough to be measured as conductivity changes.
Figure 7 also showed clearly pH differences depending upon ORP measurement sequences of phase I, phase II, and no disturbance measurements.Here, no disturbance meant that there was no ORP measurement but only pH measurement at all.All pH data of no disturbance approached to 6, and of phase I and II to 5.7 in 80 min later, respectively.Namely, there seemed to be a slight dependence of pH variation toward to 5.7 on types of measurement phases or total ORP measurement duration.One concern was the relatively vigorous deviations of pH from initial stage up to 100 min.It has already been pointed out how difficult the pH of low temperature below 10 °C could be measured because of high electrical insulation of UPW. 12 Hence, this phenomenon might be caused by many electrolysis conditions such as electrolysis current and voltage, UPW purity and temperature, ORP measurement duration, pH sensor surface, and so on.This concern was left open for further discussion.
It is well known that two extreme electrolysis reactions of Eq. 1 take place for two extreme cases of pH 0 and 14 at anode in pure water, since reactions had something to do with only H or OH related elements. 15owever, in electrolysis of UPW of pH 7, it is not reasonable to expect reactions of Eq. 1. H + and OH − ion concentrations in UPW are near 1E-7 mol l −1 each, and its resistivity is around 18.2 MΩ almost like an insulator.It is, therefore, somehow vague to understand and express the exact process of electrolysis in UPW since electrolysis must be carried out under conditions of non-equilibrium and electrical insulation.Focusing on electrolysis reactions for anode water production, it is rather probable that electrolysis could generate easily non-charged chemical species such as OH°radical.It has been well understood that OH°radical is very stable as a dilute gas but decays very rapidly in the condensed phase, which is pervasive in some situations. 16,17Gathering up all these background descriptions together with observation of very highly positive ORP phenomenon just after electrolysis as in Fig. 4, reactions in Eq. 2 of four detailed reactions may take place.One point should be stressed regarding these reactions, which was the build-up of very low concentration of H + ion in anode water.It was clearly observed that there was a slight decrease of pH in anode water converging to pH 5.7 as electrolysis currents and ORP measurements increased.Hence, with one condition that the initial UPW is a good insulator near pH 7, electrolysis reactions as in Eq. 2-1 and (2-2) can be expressed likely to take place.

→
Regarding a delicate behavior of the convergence to pH 5.7 in anode water, AEM is considered to function for accumulation of H + ions generated as in Eq. 2-2.Since AEM is considered to let the produced OH − ions in cathode drift to anode water across it while the produced H + ions in anode water to be blocked and accumulated in anode water, the reactions of AEM after electrolysis can be expressed as in Eqs.2-3 and 2-4.Therefore, there must coexist three species of chemicals of H + ions, OH − ions, and OH°radicals in anode water even under amphoteric non-equilibrium state.Then, total H + ions would contribute to pH readings and OH°radicals mostly to reactions of ORP readings in anode water.Finally, two reactions of OH − ions as in Eq. 2-5 could also take place to reduce their concentrations of H + and OH − ions to satisfy Le Chatelier's principle, as long as amphoteric non-equilibrium existed, since they did not re-store equilibrium states quickly enough due to the  Actually, it was expected that there coexisted five chemical species in anode water around pH 6 during electrolysis and ORP measurement, which are H, H + , OH − , O 3 and OH°.They can be classified functionally as follows.H + ions contribute to pH readings, OH − ions to weak negative ORP readings, and O 3 and OH°radicals to strong positive ORP ones.Also, H and O 3 species can be considered insignificant in a viewpoint of time-dependent ORP measurement since they are expected to be gasified mostly away quickly by themselves and also not to affect conductivity change.OH°radicals develop positive ORP by producing OH − ions, which affects indispensably to conductivity because of negative ion formation.Namely, unreacted OH°radical species decreased and produced OH − ions species increased as ORP measurement time increased.Hence, ORP differences between phase I and II conditions must indicate the concentration differences of unreacted OH°F ECS Advances, 2023 2 040510 radicals or produced OH − ions just after each ORP measurement point as shown in Fig. 5.It was clearly observed that conductivity increased as total ORP measurement time increased because the total ion concentrations of produced OH − ions increased, and that results of conductivity measurements for phase I and II were different.Therefore, it was concluded that ORP decrease and conductivity increase as measurement time elapsed were caused by OH − ions increase and OH°radicals decrease.The final comment was that the difference of two ORP and conductivity readings in phase I and II conditions were caused not by O 3 but only by OH°radicals.O 3 could affect to ORP readings but could not to conductivity as mentioned above, which was also very import to ignore O 3 contribution to ORP measurement.One, though, must be stressed that pH was not basically changed much all the time of measurements because pH measurement mechanism was dependent purely only on H + ions and had nothing to do with appearances of other chemical species theoretically.This fact was proven by confirming the convergence of pH to 5.7.
Reviewing the above discussion, once OH°radical was once formed and existed intact in anode water, it maintained its form in the electrochemically insulated condition at near pH 6 until it met new chemically reactive environment such as ORP measurement inducing galvanic circuit.As soon as Pt electrode was inserted in anode water for ORP measurement, electrons should be generated from Ag reference electrode and then induced to flow to Pt electrode because of strong reduction tendency of OH°radical transforming to OH − ions by gaining electrons.Namely, an OH°radical kept oxidizing an Ag atom that was losing an electron, which developed the positive oxidation potential readings between Pt in anode water and Ag reference electrodes.These reactions were expressed as  Table I displays reduction potentials of some important reactions, and it is shown that OH°radical is even higher than ozone. 18,19As well known, OH°radical is a so strong oxidizer that it reduces itself to the form of OH − ion by oxidizing Ag of reference cell to Ag + at the same time.However, as summarized by Huang et al., 20 there is no direct and accurate method to measure OH°radical directly so far.This condition creates even more difficulty in measuring it in pure anode water of pH 6.As mentioned at Introduction, there appear very various material surfaces to be cleaned at the same time in front and backend processes of EUV semiconductor technology such as metals, compounds, and low and high k dielectrics of various chemical forms of hydrides, hydrates, oxides, nitrides, carbides and so on. 21,22Therefore, the wet cleaning requirements now become more stringent in view of EUV cleaning and rinsing concepts as well as of oxide formation before the following process step of nano-scale device fabrication.It has already been proven that cathode water showed the excellent performance for commercial application to clean the contaminated Sn on EUV mask.The very interesting mechanism of cathode water cleaning was proposed as SOLUS (selective oxidation and lift-up of scums). 23In this work, it was observed that anode water was even stronger for oxidation without involving oxygen species than cathode water.Also, OH°radical was confirmed to transform itself to OH − through ORP measurement itself.The existence of OH − is well known to be prerequisite for particle removal in conventional cleaning mechanism.These characteristic phenomena of strong oxidation without oxygen species together with OH − self-formation would oxidize nano-particles selectively, lift-up, and remove them from semiconductor surfaces sequentially or metal contaminants to be ionized away.Therefore, anode water can be applied to EUV semiconductor rinsing and cleaning where no chemicals as well as no oxygen species should appear, because it is a reservoir of OH°radical for OH − ion formation necessary to clean EUV material surfaces without forming nano scale oxide layer.Anode water is even able to apply to hydrophobic surface by enhancing wettability because of providing strong oxidation environment with no oxygen species.Extra positive view point regarding applicability of anode water to EUV semiconductor or other surface treatments is that it can be reserved stable for certain time periods in insulating materials such as Teflon or glass.A final comment on these observations was that all reactions took place at near pH 6 of anode water and total ions involved were very low in concentration.These reactions were not easily measured and understood, but the observed phenomena were clearly delivering what's happening in anode water because anode water was so pure and relatively insulating.

Conclusions
Highly oxidative anode water using AEM was produced by electrolysis.Measurements of ORP, pH, and conductivity were carried out together and very significant understandings were extracted about ultra-pure anode water of UPW domain for the first time.The convergence to pH 5.7 of anode water were mainly due to effect of AEM.OH°radicals were reduced to OH − ions concurrently oxidizing Ag atoms of reference cell to Ag + ions during ORP measurements.The dominant chemical species in anode water contributing to positive ORP readings was concluded to be mostly OH°radicals.The differences of ORP and conductivity readings between shorter and longer ORP measurement times were caused by concentration difference of OH°radicals conversing to OH − ions.H + ions contributed to pH readings, OH°radicals to ORP, and OH − ions to conductivity independently.It should be pointed out that OH°r adicals in anode water electrolyzed with AEM would provide very strong oxidative condition under no oxygen appearance for long time as far as being in intact, which is essential for EUV semiconductor applicability for cleaning and rinsing.Hence, anode water can be applied for EUV semiconductor rinsing and cleaning where no oxygen species should appear for preventing oxide layer formation during cleaning, or can even remove the native oxide developed unintentionally before cleaning.

Figure 3 .
Figure 3. Electrostatic discharging effects on ORP of electrolyzed waters with CEM.

igure 5 .
Properties of electrolyzed water as ORP measurement time.(a) ORP of cathode water electrolyzed with CEM.(b) ORP and conductivity of anode water electrolyzed with AEM.

Figure 6 .
Figure 6.ORP and pH variations of anode water with AEM as for phase I.

Figure 7 .
Figure 7. pH measurements of anode waters with AEM in phase I and II, and no disturbance measurements.

Table I .
Reduction potentials of some chemical species.