Process for preparing electronic grade hydrogen peroxide

Abstract

The present application relates to the technical field of electronic grade chemical preparation, and discloses an electronic grade hydrogen peroxide preparation process, comprising: performing transient oxidation on hydrogen peroxide raw material liquid containing transition metal impurities, and converting the transition metal impurities into high valence state with inhibited catalytic activity; regulating the pH value of the solution in a short time to lock the high valence state and inhibit the valence state rotation; introducing the treated liquid into a cation exchange resin for deep purification, and ensuring the stability of the process through self-adaptive closed-loop control rules; through the synergistic sequence of chemical regulation and then physical separation, the present application blocks the decomposition path dependent on the valence cycle of metal ions, so that the final product obtains an inherent stability independent of the limit purity, and the quality reliability of the product in a complex application environment is improved.

Description

Technical Field

[0001] This invention relates to a process for preparing electronic-grade hydrogen peroxide, belonging to the field of electronic-grade chemical preparation technology. Background Technology

[0002] Currently, particularly for high-purity hydrogen peroxide used in semiconductor wet processes, the core of technological development has long focused on how to deeply remove metallic impurities from raw materials, especially catalytically active transition metal ions. The industry generally adopts a combination of purification methods such as distillation, multi-stage ion exchange, and adsorption to reduce the physical concentration of impurities to extremely low levels, such as trillions of parts per billion. This has become an industry consensus for measuring product quality and process level. The reason is that the decomposition of hydrogen peroxide is essentially a chain reaction catalyzed by trace metal ions. Its catalytic efficiency does not simply depend on the absolute number of ions, but on specific metal ions, such as iron ions, undergoing continuous and cyclical redox transformations between different valence states, such as ferrous iron and ferric iron. Each cycle decomposes several hydrogen peroxide molecules. Therefore, as long as this valence state cycle can be started and maintained, the decomposition reaction will continue to occur even at extremely low ion concentrations.

[0003] However, in cutting-edge semiconductor manufacturing practices, a long-standing and difficult-to-attribute phenomenon is that even when the highest precision analytical instruments confirm that the concentration of metal ions in the product is fully compliant, such high-purity hydrogen peroxide still exhibits an unpredictable accelerated decomposition trend in actual storage, long-distance transportation in the plant system, and the complex microenvironment at the point of use. This directly leads to unstable process solution concentration and the generation of microbubbles, posing a potential risk to the uniformity of wafer surface treatment. Faced with this dilemma, the intuitive improvement approach for those skilled in the art is usually to continue seeking raw materials with higher purity or to develop ion exchange resins with greater adsorption capacity and stronger selectivity. However, this path quickly reaches the physical limits and cost-effectiveness boundaries in reality. More importantly, it still remains within the traditional framework of reducing the number of impurities and does not address the core of the problem: as long as there are metal ions in the system that can undergo valence state cycling, this catalytic decomposition engine has not truly been shut down.

[0004] Specifically, existing technologies have the following shortcomings: 1. Existing purification methods focus solely on removing impurities, lacking effective means to inhibit the activity of residual metal ions, even those present in extremely small quantities, that still possess catalytic activity; 2. Simply pursuing a reduction in impurity concentration ignores the dynamic stability requirements of hydrogen peroxide throughout the entire application chain, failing to fundamentally prevent its decomposition under complex operating conditions. Therefore, how to inhibit the catalytic activity of trace residual metal impurities in hydrogen peroxide and block their valence state cycling pathways from the reaction mechanism level without excessively increasing purification costs has become the technical problem to be solved by this invention. Summary of the Invention

[0005] This invention provides a process for preparing electronic-grade hydrogen peroxide, the main purpose of which is to solve the problem that existing technologies only focus on reducing impurity concentration, while lacking means to inhibit the catalytic activity of residual trace impurities, resulting in insufficient stability of the product in practical applications.

[0006] To achieve the above objectives, the present invention provides a process for preparing electronic-grade hydrogen peroxide, comprising:

[0007] Step 1: Ozone and hydrogen peroxide feedstock containing transition metal impurities are reacted in a gas-liquid contact unit for a contact time of 0.5 to 5 seconds to oxidize the transition metal impurities to a higher valence state.

[0008] Step 2: Within 30 seconds after the gas-liquid reaction is completed, adjust the pH value of the hydrogen peroxide feed solution to the range of 3.0 to 4.5 and maintain it.

[0009] Step 3: The hydrogen peroxide feed solution treated in Step 2 is introduced into the cation exchange resin for purification; and in the process of Step 1, a closed-loop control rule is also followed, which includes: monitoring the ozone concentration at the exhaust gas outlet of the gas-liquid contact unit to determine the ozone dissipation rate, and adjusting the operating power of the ozone generator used to generate ozone based on the rule of comparing the net reaction dissipation rate after compensating the background dissipation rate determined according to Step 4 with the actual ozone supply verified according to Step 5.

[0010] Step 4: According to the set time frequency, periodically reduce the ozone flow rate of the hydrogen peroxide feed solution to the baseline value and maintain it for the set calibration time. The ozone dissipation rate monitored and recorded during this calibration time is determined as the background dissipation rate corresponding to the current process temperature.

[0011] Step 5: Establish a benchmark correlation model between a set of electrical characteristic parameters of the ozone generator and its ozone output, and monitor the electrical characteristic parameters of the ozone generator during process operation. Substitute the monitored values ​​into the benchmark correlation model to calculate the actual ozone supply.

[0012] Preferably, the goal of the closed-loop control rule is to adjust the operating power of the ozone generator so that the ratio of net reaction dissipation rate to actual ozone supply is not less than 0.95 when the flow rate of hydrogen peroxide feedstock and the concentration of transition metal impurities change.

[0013] Preferably, the pH range is maintained for 2 to 10 minutes in step two.

[0014] Preferably, the ozone in step one is generated in situ.

[0015] Preferably, the gas-liquid contact unit in step one uses a Venturi injector or a static mixer.

[0016] Preferably, the cation exchange resin in step three is a strongly acidic cation exchange resin.

[0017] Preferably, the transition metal impurities include iron ions.

[0018] Preferably, in step two, the pH value of the hydrogen peroxide feed solution is adjusted to the range of 3.0 to 4.5 using ultrapure acid or ultrapure alkali through online pH monitoring and an automatic acid-base addition system.

[0019] Preferably, when the deviation between the actual ozone supply calculated in step five and the command supply received by the ozone generator continues to exceed the health status threshold, the closed-loop control rule, in addition to compensating for the operating power, also generates and outputs a predictive maintenance alarm.

[0020] Preferably, after step three, a physical purification step is also included, which specifically involves vacuum degassing the hydrogen peroxide feedstock solution through a membrane contactor to remove dissolved gases.

[0021] Compared with the prior art, the beneficial effects of the present invention are:

[0022] 1. This invention provides a method for establishing the intrinsic stability of hydrogen peroxide. First, a transient oxidative shock is used to uniformly convert catalytically active transition metal impurities in the feed solution into high-valence states with suppressed catalytic activity. Then, the pH value of the liquid environment is immediately controlled and maintained within a specific acidic range. This is not conventional acidic storage, but rather utilizes the reaction kinetics of this environment to inhibit the reduction of the formed high-valence metal impurities back to catalytically active low-valence states. The close connection between these two steps blocks the decomposition path dependent on the valence state cycle of metal ions within the specific compound system of hydrogen peroxide. This makes the stability of hydrogen peroxide no longer dependent on the limit of subsequent impurity removal, but rather stems from the fact that it no longer possesses the intrinsic conditions for catalytic decomposition reactions to occur.

[0023] 2. Based on the oxidation and locking of the valence state of the metal impurities, the treated hydrogen peroxide feed solution is then introduced into the cation exchange resin. Since the target metal impurities entering the resin have been pre-converted into a higher valence state form with greater affinity to the resin, and their catalytic activity is suppressed, the phenomenon of gas blockage caused by the catalytic decomposition of hydrogen peroxide inside the resin bed is avoided. Therefore, this process sequence of chemical regulation followed by physical separation improves the exchange efficiency and service life of the cation exchange resin, which serves as the standard purification unit, for the target impurities. The upstream and downstream processes of the entire purification process thus form an effective technical synergy.

[0024] 3. This invention also provides a method for dynamically controlling the hydrogen peroxide purification process. While performing instantaneous oxidation shock, it monitors the concentration of unreacted ozone at the exhaust gas outlet of the gas-liquid contact unit and compares this concentration with the initial amount of ozone supplied. The obtained ozone dissipation rate can directly characterize the real-time load of impurities in the feed liquid. This dissipation rate is then used to adjust the amount of ozone generated. Thus, ozone, as a chemical reactant, also becomes a carrier of process information. The entire system establishes a self-adjusting operating loop based on actual reaction requirements without adding additional invasive liquid analysis components to cope with the dynamic fluctuations in the concentration of impurities in the upstream feed liquid.

[0025] 4. Based on the above adaptive adjustment loop, this control method can also periodically supply ozone at a preset baseline concentration. The ozone dissipation rate monitored at this time mainly reflects the background dissipation caused by factors such as ozone decomposition itself at the current process temperature. The system uses this background dissipation rate as a dynamic benchmark to compensate for and correct the total dissipation rate monitored during normal operation. By actively introducing a calibration period into the time series, the system can distinguish between impurity reaction dissipation and temperature-induced dissipation, so that the adaptive adjustment of ozone supply is independent of the dependence on process temperature changes, thereby improving the operational stability of the entire process under different environments and operating conditions. Attached Figure Description

[0026] Figure 1 This is a timing diagram of the pH adjustment and valence state locking process of the present invention;

[0027] Figure 2 This is a graph showing the dynamic response of the adaptive process control system of the present invention.

[0028] Figure 3 This is a functional block diagram of the adaptive process control system of the present invention. Detailed Implementation

[0029] To make the objectives, technical solutions, and advantages of this invention clearer, the invention will be described in further detail below. Obviously, the described embodiments are only some embodiments of this invention, not all embodiments. All other embodiments obtained by those skilled in the art based on the embodiments of this invention without creative effort are within the scope of protection of this invention.

[0030] The electronic-grade hydrogen peroxide preparation process disclosed in this invention is mainly composed of a catalytic activity inhibition module, a deep purification module, and an adaptive process control system that runs throughout the process. The catalytic activity inhibition module is responsible for pretreating the chemical activity of transition metal impurities in the raw material solution, the deep purification module is responsible for physically removing the pretreated impurities, and the adaptive process control system ensures that the entire process, especially the catalytic activity inhibition step, can be adjusted according to the dynamic changes in the operating conditions. The three modules are sequentially connected in the process flow and work synergistically in the functional logic to jointly establish and ensure the stability of electronic-grade hydrogen peroxide.

[0031] In semiconductor wet processes, even hydrogen peroxide solutions with adequate physical metal ion concentrations can still decompose in practical applications due to trace residues and the catalytic activity of metal ions with valence state cycling capabilities. To address this challenge, this process incorporates a catalytic activity inhibition step before deep physical purification. The core of this step lies in disrupting the catalytic cycle pathway of metal ions at the level of chemical reaction kinetics through a transient oxidation shock and a rapid pH lock. Specifically, this step involves first reacting a hydrogen peroxide feed solution containing transition metal impurities, such as iron ions, with ozone in a gas-liquid contact unit. The ozone is preferably generated in situ by corona discharge of high-purity oxygen to avoid introducing external impurities. The gas-liquid contact unit can be a venturi injector or a static mixer, among other efficient mass transfer devices, to ensure thorough mixing and reaction of the gas and liquid phases within a short time. The contact time is set within a specific process window, from 0.5 to 5 seconds. This window is based on the fact that when the contact time is less than 0.5 seconds, the ozone reacts with low-valence metal ions (such as iron ions) in the feed solution. The redox reaction may be incomplete, resulting in some catalytically active centers failing to transform. When the contact time exceeds 5 seconds, the oxidation efficiency for the target ion is no longer significantly improved, and it may increase the trace decomposition effect of ozone on hydrogen peroxide itself. Therefore, this range is obtained after a technical trade-off between ensuring oxidation efficiency and controlling side reactions. Through this operation, catalytically active transition metal impurities in the feed solution are oxidized to a stable high valence state where their catalytic activity is inhibited (e.g., ...). The core mechanism of the transition metal impurities in this invention, which catalyze the decomposition of hydrogen peroxide, lies in the cyclic transformation of the metal ions between different valence states. Taking iron ions (Fe) as an example, divalent iron ions (Fe2+) It can reduce hydrogen peroxide and oxidize itself to ferric ions (Fe3+). );and It can also oxidize another molecule of hydrogen peroxide, and then reduce itself back to its original state. This leads to a catalytic cycle, and this catalytic mechanism also applies to other trace transition metals with multiple stable oxidation states, such as copper ions (…). ) and manganese ions ( These are also potent catalysts for the decomposition of hydrogen peroxide. This process, through a transient oxidation step using ozone, aims to uniformly convert these metal ions capable of valence cycling into stable, high-valence states whose catalytic activity is suppressed (such as...). , , (etc.), thereby universally blocking such catalytic decomposition pathways; to prevent the reduction of already formed high-valence ions in subsequent processes, within 30 seconds after the gas-liquid reaction ends, the system uses online pH monitoring and an automatic acid-base addition system to rapidly adjust and maintain the pH value of the hydrogen peroxide feed solution to the range of 3.0 to 4.5 using ultrapure acid or ultrapure alkali. In the pH adjustment process described in step two, the primary principle for the selected ultrapure acid or ultrapure alkali is to avoid introducing new metal ions or impurity ions that are harmful to semiconductor processes. Therefore, the preferred acid regulator is high-purity sulfuric acid ( ) or nitric acid ( When a pH increase is required, high-purity ammonia is the preferred alkaline adjuster. Instead of strong bases such as sodium hydroxide (NaOH) or potassium hydroxide (KOH), it fundamentally prevents the introduction of alkali metal ions. The risk of pH range is as follows: when the pH value is higher than 4.5, It is prone to hydrolysis, forming hydroxide colloids or precipitates, which pose a risk of physical blockage to subsequent ion exchange processes. Furthermore, when the pH value is below 3.0, the excessively acidic environment may adversely affect the performance of subsequent ion exchange resins. Therefore, a pH range of 3.0 to 4.5 is specified to inhibit [hydrolysis]. Towards The rotation, while also being compatible with downstream processes, allows for a working window. This locked-in valence state is preferably maintained for 2 to 10 minutes to ensure the establishment of chemical equilibrium within the system. In this way, through the close connection between oxidation and locking, the catalytic activity of trace metal impurities is blocked.

[0032] After the chemical activity of metal impurities is inhibited, these passivated impurities need to be physically removed from the system. Therefore, this process introduces the hydrogen peroxide feed solution treated in step two above into a cation exchange resin for deep purification. In this design, a strongly acidic cation exchange resin is preferred because this type of resin is effective for high-valence metal ions, especially... It showed more than Higher exchange selectivity and affinity, and because the catalytic activity of metal impurities entering the resin bed is suppressed in advance, the phenomenon of catalytic decomposition of hydrogen peroxide to produce oxygen inside the resin bed is avoided, which would cause gas blockage or physical impact damage to the resin balls. This process sequence of chemical regulation followed by physical separation improves the operating efficiency and service life of the ion exchange unit compared to traditional processes, forming a technical synergy between upstream and downstream processes.

[0033] However, in industrial production environments, the concentration of impurities in upstream feed liquids can fluctuate dynamically and unpredictably. Using a fixed dose of ozone for oxidation carries the risk of incomplete oxidation when impurities exceed limits, or ozone waste when impurity levels are low. To address this dynamic adaptability issue, this process also follows an adaptive closed-loop control rule during the oxidation step. This rule requires establishing a benchmark correlation model to determine the actual output of the ozone generator. The determination procedure is as follows: During system commissioning, without introducing hydrogen peroxide feed liquid, the ozone generator is driven from its lowest power to its highest power, and its electrical characteristic parameters (voltage, current, power) at each power level are simultaneously recorded along with the actual ozone output measured by the UV ozone analyzer at the exhaust gas outlet. By correlating and fitting these two sets of data, a mapping model from electrical characteristic parameters to ozone output can be obtained. This model is embedded in the process controller. During normal process operation, the controller monitors the electrical characteristic parameters of the ozone generator in real time and substitutes the monitored values ​​into the benchmark correlation model to calculate the verified actual ozone supply. This change transforms the controller's perception of ozone supply from indirect command values ​​to a verified physical quantity based on the actual state of the equipment; simultaneously, the controller monitors the ozone concentration at the exhaust outlet of the gas-liquid contact unit in real time. The total ozone dissipation rate was calculated based on mass balance. This total dissipation rate includes not only the effective dissipation for oxidizing metal impurities, but also the background dissipation caused by factors such as ozone decomposition at a specific process temperature. To distinguish between the two, the controller is also configured to perform online calibration of the background dissipation rate at a set time frequency, such as every 10 minutes. During this calibration period, the controller reduces the supplied ozone flow rate to a preset baseline value, which is low enough that its reaction with impurities is negligible. The ozone dissipation rate monitored and recorded at this time is then determined as the background dissipation rate corresponding to the current process temperature. Finally, the controller adjusts according to the core algorithm, which first subtracts the newly calibrated background dissipation rate from the total dissipation rate to obtain the net reaction dissipation rate that is only related to impurity reactions. Subsequently, the controller adjusts the operating power of the ozone generator to achieve a ratio of net reactive dissipation to actual ozone supply. The ozone content is stably maintained at a preset target value, such as not less than 0.95. Through this closed-loop control rule, the system utilizes ozone as a reactant to simultaneously achieve online, non-invasive monitoring of impurity load, dynamic compensation for the influence of environmental factors such as temperature, and calibration of actuator performance drift, thereby ensuring the accuracy and robustness of the oxidation step. As a numerical example, assuming that at a certain moment, the system calculates the actual ozone supply based on the electrical fingerprint of the ozone generator. The concentration of unreacted ozone measured at the exhaust gas outlet is 100 units. For a concentration of 5 units, the total dissipation rate is... With a concentration of 95 units, if the background dissipation rate obtained from the latest system calibration is... For a concentration of 10 units, the net reaction dissipation rate is... At a concentration of 85 units, the dissipation ratio at this moment is If this value is below the target threshold of 0.95, it indicates that the ozone supply is relatively high compared to the current impurity load. The controller will correspondingly reduce the operating power of the ozone generator until the ratio rises back above 0.95. Furthermore, when the deviation between the actual ozone supply and the supply commanded by the controller continuously exceeds the preset health status threshold, the closed-loop control rule, in addition to power compensation, will generate and output a predictive maintenance alarm to alert operators to the performance degradation trend of the ozone generator. The adaptability of this process is reflected not only in the types of impurities but also in its robustness to impurity concentration fluctuations. The adaptive closed-loop control rule calculates the net reaction dissipation rate (Δ) in real time. ) and actual ozone supply ( The system dynamically adjusts the ozone generator's operating power based on the ratio of the raw material's transition metal impurity concentration to its value. This means that when the concentration of transition metal impurities in the raw material is low, the system will automatically reduce the ozone supply, saving energy while ensuring oxidation effect. When fluctuations occur in the upstream raw material, causing a sudden increase in impurity concentration, the closed-loop control system can quickly increase the ozone generator's power in a very short time, ensuring sufficient oxidant to oxidize the excessive impurities to a high valence state, thereby ensuring the continuity and stability of the final product quality.

[0034] Finally, considering that ozone injection and its subsequent decomposition in this process can lead to a supersaturated state of dissolved oxygen and other inert gases in the final product, this could pose a physical risk to the uniformity of wafer surface treatment due to microbubble formation in subsequent applications. Therefore, this process can also include a physical purification step after the deep purification step. Specifically, this step involves vacuum degassing the hydrogen peroxide solution through a membrane contactor composed of a hydrophobic porous hollow fiber membrane to remove dissolved gases. Since this process has improved the chemical stability of hydrogen peroxide, making it independent of the traditional dissolved oxygen environment to inhibit decomposition, it can undergo deep degassing. This not only reduces the potential risk of microbubbles but also synergistically removes any trace volatile organic compounds that may remain in the system, thus ensuring the quality of the final product from both chemical and physical perspectives.

[0035] Example 1: In the advanced node wet cleaning process of a large-scale semiconductor manufacturing plant, the production line faces yield fluctuation problems, which manifest as microbubble defects on the wafer surface after treatment with hydrogen peroxide-based cleaning solution. The factory's incoming inspection of all batches of electronic-grade hydrogen peroxide raw materials shows that the physical concentration of transition metal impurities, especially iron ion concentration, is consistently maintained at the trillions of parts per billion level, which meets the requirements of the technical specifications. The factory's original purification process relies on multi-stage strong acid cation exchange resin for purification. Its design idea is to reduce the physical quantity of impurities, but this method has failed to avoid the above-mentioned occasional product decomposition and microbubble generation phenomena.

[0036] To address this operating condition, the plant deployed the preparation process disclosed in the aforementioned specific embodiments into its central chemical supply unit. When a new batch of hydrogen peroxide feedstock enters the system, the process is automatically triggered. The feedstock is first introduced into a gas-liquid contact unit consisting of a Venturi injector, where it undergoes a 2-second contact reaction with the ozone generated in situ. Trace amounts of residual, still catalytically active, ferrous ions remain in the feedstock. In this step, it is oxidized to ferric ions, whose catalytic activity is inhibited. Immediately following, within 5 seconds of the reaction solution leaving the Venturi injector, the online pH monitoring and automatic acidification system is activated to adjust and maintain the pH of the liquid to 3.5. This step utilizes the reaction kinetics under this specific acidic environment to convert the already formed ferric ions into ferrous ions. Locked in a stable hydrated ion state, inhibiting the reduction of hydrogen peroxide back to ferrous ions. The path.

[0037] The hydrogen peroxide liquid treated with the above-mentioned catalytic activity inhibition step is then introduced into a downstream strongly acidic cation exchange resin bed for final physical purification. In this step, a synergistic relationship is formed between two technical features: First, because the target impurity has been pre-converted into a form with a greater affinity for the resin exchange sites. The morphology of the ion exchange unit makes its iron ion capture efficiency and total exchange capacity higher than that of directly treating iron containing other ions. / The efficiency of mixed-valence liquids is improved; at the same time, since the catalytic activity of metal ions entering the resin bed is inhibited, the side reaction of catalytic decomposition of hydrogen peroxide to produce oxygen inside the resin bed is avoided. This avoids operational risks such as uneven liquid flow caused by gas blockage inside the resin bed and physical wear of resin particles caused by bubble impact. This sequential combination of chemical pretreatment and physical deep removal improves the efficiency of individual links and ensures the long-term stability of the entire purification system.

[0038] By implementing this process, the original focus on controlling only the physical concentration of impurities has been transformed into simultaneously controlling the physical concentration and chemical activity of impurities. It no longer directly addresses the problem of how to avoid the catalytic decomposition of trace impurities in complex plant pipelines, but rather breaks the chemical conditions for impurities to cycle in their valence states at the source, enabling the hydrogen peroxide liquid itself to acquire a stability independent of the subsequent application environment. After three months of continuous production using this process, no yield abnormalities caused by microbubbles were recorded in the wet cleaning process of the semiconductor plant, and the stability and predictability of its production process were improved.

[0039] Example 2: To objectively verify the effect of the process of the present invention on the stability of hydrogen peroxide, a comparative experiment was designed and conducted. This experiment aimed to quantitatively compare the thermal stability of the product treated by the process of the present invention with that treated only by existing deep purification technologies, under the same metal impurity concentration level. The starting material used in the experiment was the same batch of industrial-grade hydrogen peroxide with an initial concentration of 31.2%. The total iron ion concentration was measured to be 98.5 ppb using inductively coupled plasma mass spectrometry (ICP-MS, with a detection limit of not less than 1 ppt). Two parallel purification lines were set up in the experiment: a control group and the sample group of the present invention. Both lines used the same specification and batch of strongly acidic cation exchange resin. The process flow configuration of the control group was as follows: The existing deep purification method involves directly passing hydrogen peroxide feedstock through a strongly acidic cation exchange resin column at a set flow rate. The process flow of the sample group of this invention follows the technical solution disclosed in the aforementioned specific embodiments. That is, before entering the same strongly acidic cation exchange resin column, a catalytic activity inhibition step is first performed. The key operating parameters of this step are set as follows: a static mixer is used as the gas-liquid contact unit, the contact time between ozone and hydrogen peroxide feedstock is controlled at 1.5 seconds, after the gas-liquid reaction is completed, the pH value of the liquid is adjusted and maintained at 4.0 within 10 seconds, and this state is maintained for 5 minutes before the liquid is introduced into the ion exchange resin column. Both sets of experiments are carried out under the same ambient temperature and liquid flow rate to eliminate interference from other variables.

[0040] After the two processing lines were running stably, the final purified products were collected from their outlets for analysis. The results showed that the two groups of products were on the same order of magnitude in terms of the physical removal effect of metal impurities. The total iron ion concentration of the control group product was 8.1 ppt, while that of the product in this invention group was 7.5 ppt. Subsequently, samples of the two groups of products, each with an initial concentration of 31.1%, were taken and placed in clean PFA sealed containers, and incubated at 85°C. Accelerated aging tests were conducted by continuous heating in a constant temperature environment for 24 hours. After the test, the hydrogen peroxide concentration in the samples was measured, and the results were recorded as follows: The hydrogen peroxide concentration of the control group decreased from 31.1% to 28.3%, with a concentration decay rate of 9.0%, and the calculated decomposition rate was 0.375% / h; in contrast, the hydrogen peroxide concentration of the sample group of the present invention decreased only from 31.1% to 30.7%, with a concentration decay rate of 1.3%, and the corresponding decomposition rate was 0.054% / h. The test results show that, under similar total iron ion concentrations, the thermal decomposition rate of the sample treated by the process of the present invention is much lower than that of the sample treated only by ion exchange. This difference indicates that in the sample group of the present invention, trace residual iron ions have been converted into a high-valence state with suppressed catalytic activity due to the treatment of the catalytic activity inhibition step, thereby blocking their catalytic decomposition pathway under high-temperature conditions. The data from this test confirms that by using the process of the present invention, while achieving deep physical removal of transition metal impurities, the catalytic activity of trace residual impurities is also inhibited by chemical means, and the obtained hydrogen peroxide product has higher stability.

[0041] To verify the effectiveness of the physical purification step, a sample of the present invention was taken as sample A after treatment with a strongly acidic cation exchange resin and before vacuum degassing. This sample was then degassed using a membrane contactor and taken as sample B. Both samples were analyzed using an online dissolved oxygen analyzer with a range of 0-100 ppm and an accuracy of ±0.1 ppm. The dissolved oxygen concentration in sample A was measured to be 65.3 ppm, indicating a supersaturated state, while the dissolved oxygen concentration in sample B decreased to 4.8 ppm. Furthermore, samples A and B were placed in two identical reduced-pressure environments. Using a laser particle counter, it was observed that sample A produced a large number of microbubbles after the pressure decreased, while no significant microbubbles were observed in sample B under the same conditions. This supplementary experimental data indicates that the vacuum degassing step can effectively remove excess dissolved gas introduced into the solution by the ozone process, thereby reducing the risk of microbubble formation due to pressure or temperature changes in subsequent applications.

[0042] Example 3: This example combines Figures 1 to 3 The implementation process of an electronic-grade hydrogen peroxide preparation method is described, such as... Figure 1As shown, the flowchart illustrates the pH adjustment and valence state locking process of this invention. After the oxidized liquid enters from the gas-liquid contact unit, the online pH monitoring module monitors the pH value in real time and feeds back the pH data to the automatic control system. The system then determines the pH adjustment requirement. If the pH value is determined to be too high, the ultrapure acid supply is activated to add an acidic regulator; if the pH value is determined to be too low, the ultrapure alkali supply is activated to add an alkaline regulator. This rapid adjustment operation within 30 seconds adjusts the pH value of the liquid to the range of 3.0-4.5 and maintains a stable pH state. This state lasts for 2-10 minutes to complete the valence state locking, thus... After being stably locked in place, the treated liquid is finally transferred to the ion exchange resin unit via a buffer storage stage.

[0043] like Figure 2 As shown in the figure, the horizontal axis represents time (seconds), the left vertical axis represents ozone generator power (%), and the right vertical axis represents the net reaction dissipation ratio. The net reaction dissipation ratio, represented by the dashed triangle, drops sharply at approximately 20 seconds, indicating a sudden increase in impurity load. In response, the ozone generator power, represented by the solid circle, begins to climb rapidly from the baseline level of 25% at approximately 30 seconds, reaching 100% at 60 seconds. As the power increases, the net reaction dissipation ratio also recovers, returning to above the target threshold of 0.95 at approximately 50 seconds, demonstrating the system's adaptive adjustment capability.

[0044] like Figure 3 As shown, the ozone generator produces ozone according to the operating power command issued by the process controller and supplies it to the Venturi injector or static mixer, which serves as the gas-liquid contact unit. At the same time, the electrical characteristic parameter monitoring module collects the electrical fingerprint of the generator and feeds it back to the controller. The reference correlation model in the controller calculates the actual ozone supply based on this fingerprint. At the other end of the reaction, the exhaust gas ozone concentration monitoring module measures the ozone concentration at the exhaust gas outlet. Concentration, combined with a dynamic benchmark periodically determined by the online background dissipation rate calibration module. This allows the controller to determine the net reaction dissipation rate through the dissipation rate calculation module. Ultimately, the closed-loop control rules within the controller adjust the operating power command to achieve the ratio In addition, when the deviation between the instruction and the actual supply exceeds a threshold, the system will also generate a predictive maintenance alarm.

[0045] Example 4: In the initial deployment and debugging stage of the process of the present invention, in order for its adaptive process control system to reflect the performance characteristics of specific physical equipment and make accurate responses under changing operating conditions, a systematic engineering calibration procedure is required. This procedure is started in an offline controlled initial state. Its target is the purification system to be put into production. The system has completed physical installation. The power supply circuit of its ozone generator integrates electrical sensors that can measure voltage, current and power. An ultraviolet ozone analyzer has also been installed at the exhaust gas outlet of its gas-liquid contact unit. Before the procedure is started, there is no hydrogen peroxide raw material liquid in the system. Only high-purity oxygen with a specification of 99.99% is supplied to the ozone generator.

[0046] The first step of the procedure is to establish a benchmark correlation model between a set of electrical characteristic parameters of the ozone generator and its ozone output. The calibration subroutine is initiated through the process controller. This subroutine automatically adjusts the operating power of the ozone generator in a stepwise manner, starting from 10% of its rated power and increasing in 5% increments to 100%. At each power setpoint, the system runs stably for 3 minutes, and the average value of the electrical sensor readings during this period is recorded. ) and the average value of the ultraviolet ozone analyzer readings ( ), through the collected ( , By performing second-order polynomial regression fitting on the data point pairs, a mapping function that characterizes the electrical power to the actual ozone production under the specific generator's healthy state can be obtained. This function is then fixed as the baseline correlation model.

[0047] The next step in the procedure is to determine the key thresholds and parameters in the closed-loop control rules. The first step is to determine the baseline value used for background dissipation rate calibration. After establishing the baseline correlation model, ultrapure water free of transition metal impurities is introduced into the system, and an initial low ozone supply is set, corresponding to 15% of the generator's rated power output. The ozone dissipation rate at this point is recorded. Subsequently, the ozone supply is gradually increased in small increments, and the change in ozone dissipation rate is continuously monitored. When the growth slope of the ozone dissipation rate is observed to stabilize at a plateau, meaning its value no longer changes significantly with small increases in the ozone supply, the dissipation rate is considered to be at its maximum. The ozone supply at the current temperature, determined by the ozone decomposition itself, serves as the baseline for subsequent online calibration. Secondly, a target value of 0.95 is set for the ratio of net reaction dissipation to actual ozone supply. This target value is determined after balancing reaction efficiency and economy. A ratio higher than 0.95 means that most of the supplied ozone participates in the effective reaction, avoiding energy waste. However, a ratio too close to 1.0 may lead to insufficient reaction margin when the system responds to sudden increases in impurity concentration. Therefore, 0.95 is determined as an engineering setting that balances high reaction conversion rate and system dynamic response capability.

[0048] By implementing the above-mentioned systematic calibration procedure, the control model and parameters that were originally based on experience are transformed into a series of traceable values ​​based on physical measurements and engineering logic. The internal logic of the entire adaptive process control system is completely transparent, and every judgment and adjustment behavior has a reproducible engineering basis. The completion of this procedure marks the transition of the purification system from the physical installation stage to the point where its control system has the initial conditions to reliably and adaptively operate in the production environment.

[0049] Example 5: When the process of the present invention is in a continuously operating production environment, due to a process abnormality in the upstream raw material supply unit, the concentration of transition metal impurities in the hydrogen peroxide raw material entering this treatment system momentarily increases, namely, ferrous ions. When the concentration of [acid] rises from the usual 5 ppb to 50 ppb within a short period of time, the system's adaptive process control rules are automatically activated to cope with this boundary condition, due to the high concentration of [acid]. Upon entering the gas-liquid contact unit, the gas undergoes an accelerated redox reaction with ozone, leading to an increase in the concentration of unreacted ozone measured by the ultraviolet ozone analyzer at the exhaust gas outlet. It drops below its detection limit within seconds.

[0050] The process controller uses this real-time data and the actual ozone supply calculated by the baseline correlation model at that time. The total ozone dissipation rate was calculated. Rapid rise, and then, after background dissipation rate Net reaction dissipation rate after compensation This also increases significantly, leading to a rise in the ratio of net reactive dissipation to actual ozone supply. When the ozone level suddenly drops to well below the control target lower limit of 0.95, the closed-loop control rules immediately increase the ozone generator's operating power command at the maximum rate in response to this deviation, until the equipment's safe operating limit is reached. This allows the actual ozone supply to be increased to a level that matches the abnormal impurity load in a short period of time. Through this adaptive adjustment mechanism, the peak contamination is oxidized and locked before entering the subsequent purification unit, ensuring the continuity of the final product quality. At the same time, because the system detects that the operating power is continuously at a high level, it will also automatically generate and output a process abnormality alarm, prompting operators to pay attention to the quality fluctuations of upstream raw materials.

[0051] Example 6: During the scale-up process of the present invention from laboratory research and development to industrial production, in order to determine the optimal working window of key process parameters in the core catalytic activity inhibition step, a systematic parameter optimization calibration procedure is required. This procedure aims to optimize the intrinsic stability of the final product by experimentally defining the impact of two interrelated variables, ozone contact time and subsequent pH lock-in value, on the optimization objective, thereby providing a reproducible process setting basis for large-scale production.

[0052] This procedure employs a matrix experiment approach, using the same batch of industrial-grade hydrogen peroxide as in Example 2 as the uniform starting material, with the product undergoing 85% [processing / processing]. The hydrogen peroxide concentration decay rate after 24 hours of continuous heating in a constant temperature environment was used as the core performance indicator to measure its intrinsic stability, i.e., the objective function of the optimization process, aiming to find the parameter combination that minimizes this indicator. The experiment set up multiple parallel processing batches. In each batch, all process conditions, including feed liquid flow rate, reaction temperature, and subsequent purification steps through a strongly acidic cation exchange resin, were kept consistent except for ozone contact time and lock pH value. The ozone contact time was set at five levels: 0.2 seconds, 0.5 seconds, 2.0 seconds, 5.0 seconds, and 7.0 seconds. The lock pH value was also set at five levels: 2.5, 3.0, 4.0, 4.5, and 5.0. By combining these two variables, a total of 25 batches of experiments were conducted.

[0053] After summarizing and analyzing the experimental data from all batches, the following results were observed: When the ozone contact time was set to 0.2 seconds, regardless of the pH adjustment, the final product concentration decay rate remained high, similar to the control group, indicating that the oxidation reaction was incomplete. When the contact time was increased to the range of 0.5 to 5.0 seconds, and the locked pH value was set to the range of 3.0 to 4.5, the concentration decay rate of all products was stably maintained at a low level, indicating that the product's intrinsic stability reached an optimized state. However, when the contact time was further extended to 7.0 seconds, the concentration decay rate no longer decreased further, and a slight increase in total organic carbon content was detected in some samples. When the locked pH value was below 3.0 or above 4.5, even if the contact time was within the aforementioned effective range, the product stability decreased compared to the optimal range, and trace amounts of colloidal substances were observed to be generated in the sample with pH 5.0.

[0054] The results of this parameter optimization procedure provide a complete set of engineering experimental verifications for the ozone contact time of 0.5 to 5 seconds and the pH value of 3.0 to 4.5 as defined in the claims. This demonstrates that the parameter window is a technically balanced result range that can synergistically suppress the catalytic activity of transition metal impurities while avoiding adverse effects on hydrogen peroxide itself or the introduction of secondary pollution.

[0055] It will be apparent to those skilled in the art that the present invention is not limited to the details of the exemplary embodiments described above, and that the present invention can be implemented in other specific forms without departing from the spirit or essential characteristics of the present invention.

[0056] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and are not intended to limit it. Although the present invention has been described in detail with reference to preferred embodiments, those skilled in the art should understand that modifications or equivalent substitutions can be made to the technical solutions of the present invention without departing from the spirit and scope of the technical solutions of the present invention.

Claims

1. A process for preparing electronic-grade hydrogen peroxide, characterized in that, include: Step 1: Ozone and hydrogen peroxide feedstock containing transition metal impurities are reacted in a gas-liquid contact unit for a contact time of 0.5 to 5 seconds to oxidize the transition metal impurities to a higher valence state. Step 2: Within 30 seconds after the gas-liquid reaction is completed, adjust the pH value of the hydrogen peroxide feed solution to the range of 3.0 to 4.5 and maintain it. Step 3: The hydrogen peroxide feed solution treated in Step 2 is introduced into a cation exchange resin for purification; In addition, during step one, a closed-loop control rule is followed, which includes: monitoring the ozone concentration at the exhaust gas outlet of the gas-liquid contact unit to determine the ozone dissipation rate, and adjusting the operating power of the ozone generator used to generate ozone based on the rule that the net reaction dissipation rate after compensating the ozone dissipation rate according to the background dissipation rate determined according to step four is compared with the actual ozone supply verified according to step five. Step 4: According to the set time frequency, periodically reduce the ozone flow rate of the hydrogen peroxide feed solution to the baseline value and maintain it for the set calibration time. The ozone dissipation rate monitored and recorded during this calibration time is determined as the background dissipation rate corresponding to the current process temperature. Step 5: Establish a benchmark correlation model between a set of electrical characteristic parameters of the ozone generator and its ozone output, and monitor the electrical characteristic parameters of the ozone generator during process operation. Substitute the monitored values ​​into the benchmark correlation model to calculate the actual ozone supply.

2. The electronic-grade hydrogen peroxide preparation process according to claim 1, characterized in that, The goal of the closed-loop control rule is to adjust the operating power of the ozone generator so that the ratio of net reaction dissipation rate to actual ozone supply is not less than 0.95 when the flow rate of hydrogen peroxide feedstock and the concentration of transition metal impurities change.

3. The process for preparing electronic-grade hydrogen peroxide according to claim 1, characterized in that, The pH range is maintained for 2 to 10 minutes in step two.

4. The process for preparing electronic-grade hydrogen peroxide according to claim 1, characterized in that, In step one, ozone is generated in situ.

5. The process for preparing electronic-grade hydrogen peroxide according to claim 1, characterized in that, In step one, the gas-liquid contact unit uses a Venturi injector or a static mixer.

6. The process for preparing electronic-grade hydrogen peroxide according to claim 1, characterized in that, In step three, the cation exchange resin is a strongly acidic cation exchange resin.

7. The process for preparing electronic-grade hydrogen peroxide according to claim 1, characterized in that, Transition metal impurities include iron ions.

8. The process for preparing electronic-grade hydrogen peroxide according to claim 1, characterized in that, In step two, the pH value of the hydrogen peroxide feed solution is adjusted to the range of 3.0 to 4.5 using an online pH monitoring and automatic acid and alkali addition system, either ultrapure acid or ultrapure alkali.

9. The process for preparing electronic-grade hydrogen peroxide according to claim 1, characterized in that, When the deviation between the actual ozone supply calculated in step five and the command supply received by the ozone generator continues to exceed the health status threshold, the closed-loop control rule, in addition to compensating for the operating power, also generates and outputs a predictive maintenance alarm.

10. The process for preparing electronic-grade hydrogen peroxide according to claim 1, characterized in that, Following step three, a physical purification step is also included, which specifically involves vacuum degassing the hydrogen peroxide feedstock solution through a membrane contactor.

Images