A DFB laser quantum well active region lateral sidewall passivation etching process

Abstract

The application discloses a DFB laser quantum well active region lateral sidewall passivation etching process. Main etching gas is introduced into an etching chamber to perform vertical direction main etching; in-process real-time acquisition of in-situ monitoring parameters is performed, a trigger signal is generated when the parameters show characteristic changes representing the start of lateral drilling etching; in response to the trigger signal, the main etching gas is maintained to be introduced, at the same time, a sidewall passivation gas is synchronously introduced and its flow rate is adjusted until the parameters reach a steady state representing the stable generation of the passivation layer; in the state, the passivation gas flow rate is dynamically adjusted based on real-time fluctuations of the parameters to maintain the suppression state; the current lateral physical size of the active region is calculated, and when the size reaches the design size and enters the natural convergence interval, it is judged that the size locking is completed; all the gases are stopped to be introduced, and the non-damage etching is completed. The application can block the lateral drilling etching from the source, realize accurate locking of the lateral size of the active region and non-damage etching.

Description

Technical Field

[0001] This invention relates to the field of DFB laser fabrication technology, and in particular to a passivation etching process for the lateral sidewalls of the active region of a DFB laser quantum well. Background Technology

[0002] As the core light source of modern optical communication networks, distributed feedback lasers (GFPS) rely heavily on the manufacturing quality of their internal quantum well active regions, which directly determines the transmission performance of optical signals and the lifespan of the devices. In manufacturing these high-precision light-emitting devices, the etching of the active region structure is an indispensable and crucial step. Traditional etching processes often focus on the material removal rate and flatness in the vertical direction, neglecting the isotropic characteristics of the etching gas during spatial diffusion. This conventional approach easily disrupts the dynamic balance between material removal and structural preservation when dealing with extremely thin quantum well layers, leaving the edges of the active region unprotected and leading to uncontrollable lateral erosion. As the main etching gas propels downwards to form ridge waveguides or bury heterojunctions, the lateral sidewalls of the quantum well are exposed to a highly corrosive environment. Due to the lack of an effective mechanism to intervene in the chemical state of the sidewalls, the etching gas inevitably laterally drills into the quantum well layers on both sides while etching downwards. This is analogous to using a high-pressure water jet to wash away mud in a ditch; the water flow not only deepens the bottom of the ditch but also continuously erodes the banks on both sides. This excessive lateral consumption directly alters the original designed lateral physical dimensions of the active region, causing the light-emitting area to narrow or even have its edges damaged, severely compromising the structural integrity of the device.

[0003] To address the aforementioned issues, some improvements have been proposed in existing technologies. A common approach is to remove the wafer after etching and immerse the etched surface in a wet chemical solution, such as sulfuric acid-hydrogen peroxide solution, to remove oxides generated during etching, followed by the formation of a sulfide protective layer on the etched surface using ammonium polysulfide solution. This offline passivation method can repair surface damage caused by etching to some extent and provide a relatively stable interface for subsequent epitaxial growth. However, this post-processing method has a fundamental flaw: it cannot address the lateral drilling that has already occurred during etching. Since the passivation step is performed only after all etching is complete, when the quantum well sidewalls are exposed to a highly corrosive plasma environment during etching, lateral drilling has already caused irreversible damage to the lateral dimensions of the active region. Subsequent chemical repair can only address the surface condition but cannot replenish the excessively consumed quantum well material. Essentially, existing technologies address how to protect the damaged surface after etching, but fail to address the core issue of how to prevent damage from occurring during the etching process. Therefore, how to precisely intervene in the chemical state of the exposed sidewalls of the quantum well during the etching process, so as to effectively block the lateral drilling of gas and strictly lock the lateral physical size of the active region, has become the key issue to achieve damage-free etching of the active region of the quantum well of the distributed feedback laser. Summary of the Invention

[0004] Therefore, the technical problem to be solved by the present invention is to overcome the defects of the quantum well sidewalls being exposed to a highly corrosive environment due to the lack of real-time chemical state intervention during the etching process of the prior art, which leads to uncontrollable lateral drilling and damage to the lateral physical size of the active region. The present invention provides a passivation etching process for the lateral sidewalls of the active region of a DFB laser quantum well, which can trigger synchronous passivation gas injection by monitoring the characteristic changes of the sidewall chemical state in situ, and dynamically maintain the sidewall suppression state based on closed-loop feedback until the lateral physical size naturally converges to the design size and then terminates the etching, thereby blocking the occurrence of lateral drilling from the source and achieving precise locking of the lateral size of the active region and non-destructive etching.

[0005] To address the aforementioned technical problems, this invention provides a passivation etching process for the lateral sidewalls of the active region of a DFB laser quantum well, comprising the following steps: The main etching gas is introduced into the etching chamber to perform vertical main etching on the active region of the quantum well of the distributed feedback laser to form a ridge waveguide structure. During the main etching process, in-situ monitoring parameters characterizing the chemical state of the quantum well sidewall are acquired in real time; when the in-situ monitoring parameters show characteristic changes that characterize the start of lateral drilling, it is determined that lateral drilling behavior has occurred on the sidewall and a trigger signal is generated. In response to the trigger signal, while maintaining the continuous flow of the main etching gas, the sidewall passivation gas is simultaneously introduced into the etching chamber, and the flow rate of the sidewall passivation gas is adjusted until the in-situ monitoring parameters reach a stable state that characterizes the stable formation of the sidewall passivation layer. Under stable conditions, the in-situ monitoring parameters continue to be acquired in real time, and the flow rate of the sidewall passivation gas is dynamically adjusted based on the real-time fluctuations of the in-situ monitoring parameters to maintain the in-situ monitoring parameters in a suppressed state that indicates that the lateral etching is effectively suppressed. The current lateral physical size of the active region is acquired or calculated in real time. When the current lateral physical size reaches the design size that represents the completion of the active region design, and the current lateral physical size enters the natural convergence interval that represents the allowable fluctuation range, it is determined that the size locking is complete. In response to the determination that size locking is complete, the main etching gas and sidewall passivation gas are stopped, and the damage-free etching of the active region of the DFB laser quantum well is completed.

[0006] In one embodiment of the present invention, the in-situ monitoring parameters include the characteristic wavelength intensity in the plasma emission spectrum that characterizes the etching byproducts of the quantum well material; the characteristic change that characterizes the start of lateral drilling is set as follows: the intensity difference of the characteristic wavelength intensity in each of the three consecutive sampling periods exceeds three times the standard deviation of the background noise of the characteristic wavelength intensity.

[0007] In one embodiment of the present invention, the characteristic wavelength includes the 451 nm spectral line corresponding to indium or the 417 nm spectral line corresponding to gallium; the sampling period is 100 milliseconds to 500 milliseconds; and the background noise standard deviation is calculated from the fluctuation value of the characteristic wavelength intensity within the first 2 seconds after etching starts.

[0008] In one embodiment of the present invention, the stable state characterizing the stable formation of the sidewall passivation layer is: the fluctuation amplitude of the in-situ monitoring parameter within five consecutive sampling periods is less than five percent of the average value of the five sampling periods; the suppression state characterizing the effective suppression of lateral etching is: the instantaneous value of the in-situ monitoring parameter is always lower than the trigger value corresponding to the characteristic change characterizing the start of lateral drilling.

[0009] In one embodiment of the present invention, the flow rate of the sidewall passivation gas is adjusted until the in-situ monitoring parameters reach a stable state characterizing the stable formation of the sidewall passivation layer. Specifically, this includes: using a proportional-integral-derivative control method, the flow rate adjustment of the sidewall passivation gas is calculated according to the following formula based on the deviation between the current value of the in-situ monitoring parameters and the target value of the stable state: ΔQ=Kp·e(t)+Ki·∫e(t)dt+Kd·de(t) / dt, where ΔQ is the flow rate adjustment of the sidewall passivation gas, e(t) is the deviation between the current value and the target value of the stable state, Kp, Ki, and Kd are the proportional coefficient, integral coefficient, and derivative coefficient, respectively, and t is time.

[0010] In one embodiment of the present invention, the flow rate of the sidewall passivation gas is dynamically adjusted based on the real-time fluctuations of the in-situ monitoring parameters to maintain the in-situ monitoring parameters in a suppressed state that indicates that lateral etching is effectively suppressed. Specifically, this includes: dividing the real-time fluctuation amplitude of the in-situ monitoring parameters into multiple levels, each level corresponding to a flow rate compensation coefficient for the sidewall passivation gas; when the real-time fluctuation amplitude rises to the upper limit of the current level, the flow rate compensation coefficient is switched to the value corresponding to the next higher level, and the flow rate of the sidewall passivation gas is increased accordingly; when the real-time fluctuation amplitude falls below the lower limit of the current level, the flow rate compensation coefficient is switched to the value corresponding to the next lower level, and the flow rate of the sidewall passivation gas is decreased accordingly.

[0011] In one embodiment of the present invention, the real-time acquisition or calculation of the current lateral physical size of the active region specifically includes: measuring the reflection spectrum of the top of the ridge waveguide of the active region in real time using an in-situ spectroscopic ellipsometer, and calculating the current lateral physical size according to the following formula based on the linear relationship between the oscillation period in the reflection spectrum and the lateral size of the active region: W=λ / (2n·Δθ), where W is the current lateral physical size, λ is the wavelength of the measured light, n is the refractive index of the active region material, and Δθ is the angular interval between two adjacent characteristic peaks in the reflection spectrum.

[0012] In one embodiment of the present invention, the current lateral physical dimension enters the natural convergence interval characterizing the allowable fluctuation range, specifically: in the measurement values ​​of the current lateral physical dimension at ten consecutive sampling times, the difference between the maximum and minimum values ​​is less than 1 nanometer, and the difference between the arithmetic mean of the ten measurement values ​​and the design dimension is less than one percent of the arithmetic mean.

[0013] In one embodiment of the present invention, the main etching gas includes a mixture of chlorine and boron trichloride, wherein the volume flow rate ratio of chlorine to boron trichloride is 1:2 to 2:1, and the ratio is pre-selected according to the material composition of the active region of the quantum well and the anisotropy of the desired etching morphology; the sidewall passivation gas includes methane or hydrogen, and when methane is used, its volume flow rate is 10% to 30% of the total volume flow rate of the main etching gas, and when hydrogen is used, its volume flow rate is 5% to 15% of the total volume flow rate of the main etching gas.

[0014] In one embodiment of the present invention, the main etching gas further includes argon, and the volume flow rate of argon is 5% to 20% of the total volume flow rate of the main etching gas; the sidewall passivation gas further includes nitrogen, which is introduced simultaneously with methane or hydrogen, and the volume flow rate of nitrogen is 20% to 40% of the total volume flow rate of the sidewall passivation gas.

[0015] The technical solution of the present invention has the following advantages compared with the prior art: The passivation etching process for the lateral sidewalls of the active region of the DFB laser quantum well described in this invention achieves precise suppression of lateral drilling behavior by monitoring the chemical state of the quantum well sidewalls in real time during the main etching process and dynamically introducing and adjusting the passivation gas based on the monitoring results. This process effectively prevents excessive erosion of the sidewalls of the quantum well active region during etching, thereby ensuring that the lateral physical dimensions of the ridge waveguide structure can accurately converge to the design value. As a result, the structural integrity of the active region of the distributed feedback laser quantum well is maintained, the optical performance and reliability of the device are improved, and the problem of difficult control of lateral dimensions in traditional processes is solved. Attached Figure Description

[0016] To make the content of this invention easier to understand, the invention will be further described in detail below with reference to specific embodiments and accompanying drawings, wherein: Figure 1 This is a flowchart of the passivation etching process for the lateral sidewall of the active region of the DFB laser quantum well according to the present invention. Figure 2 This is a flowchart illustrating the steps of adjusting the flow rate of the sidewall passivation gas according to the present invention; Figure 3 This is a flowchart illustrating the steps of dynamically adjusting flow using a tiered compensation strategy, as described in this invention. Figure 4 This is a flowchart illustrating the steps of calculating the current lateral physical dimensions of the active region according to the present invention. Detailed Implementation

[0017] The present invention will be further described below with reference to the accompanying drawings and specific embodiments, so that those skilled in the art can better understand and implement the present invention. However, the embodiments described are not intended to limit the present invention.

[0018] Reference Figure 1 As shown, this invention provides a passivation etching process for the lateral sidewalls of the active region of a DFB laser quantum well. This process does not simply shift the traditional passivation steps to the etching process; instead, it establishes a closed-loop control mechanism based on in-situ state sensing and synchronous intervention. Specifically, a main etching gas is first introduced into the etching chamber to perform vertical main etching on the active region of the quantum well to form a ridge waveguide structure. Unlike existing technologies, this process continuously acquires in-situ monitoring parameters characterizing the chemical state of the quantum well sidewalls in real time during the main etching process. These parameters reflect the actual chemical state changes of the sidewalls under the current etching environment.

[0019] When the in-situ monitoring parameters show characteristic changes indicative of the onset of lateral drilling, it means that the sidewall has begun to suffer unacceptable horizontal erosion. At this point, the process immediately determines that lateral drilling has occurred and generates a trigger signal. In response to this trigger signal, while maintaining the continuous flow of the main etching gas, a sidewall passivation gas is simultaneously introduced into the etching chamber, and its flow rate is adjusted until the in-situ monitoring parameters reach a stable state indicating the stable formation of the sidewall passivation layer. The core of this process lies in synchronization and maintenance: the introduction of the passivation gas does not interrupt the main etching process, but occurs simultaneously with it. Thus, while etching continues in the vertical direction, a protective passivation layer is generated in situ on the newly exposed quantum well sidewall, blocking the lateral diffusion path of the etching gas into the sidewall from the source.

[0020] Once the process reaches a stable state, the fluctuations in the in-situ monitoring parameters continue to be monitored in real time, and the flow rate of the passivating gas on the sidewalls is dynamically adjusted to maintain these parameters in a suppressed state, indicating that lateral etching is effectively suppressed. This is equivalent to establishing a closed-loop feedback loop: changes in the sidewall chemical state are reflected in the monitoring parameters in real time, and changes in the monitoring parameters drive adaptive adjustments to the passivating gas flow rate, thereby keeping the sidewall chemical state locked within the range where lateral etching is effectively suppressed.

[0021] Based on this, the process also acquires or calculates the current lateral physical dimensions of the active region in real time. When this dimension reaches the design dimension that represents the completion of the design and enters the natural convergence range that represents the allowable fluctuation range, it is determined that the dimension locking is complete, and then all gas is stopped, completing the non-destructive etching.

[0022] In principle, the reason why the above-mentioned technical features can solve the problems existing in the prior art is that they transform the passive mode of post-treatment in the prior art into an active mode of process intervention. The fundamental reason for lateral drilling in the prior art is that the quantum well sidewalls are in an unprotected, exposed state during the etching process, and the main etching gas itself is not selective, eroding both vertically downwards and to the sides simultaneously. This application, by introducing in-situ monitoring parameters, obtains real-time feedback on the chemical state of the sidewalls for the first time during the etching process, making the previously invisible process of when lateral drilling begins perceptible. Upon sensing the signal that drilling has started, the simultaneously introduced sidewall passivation gas can generate a protective layer as soon as the sidewalls are eroded, rather than waiting until the etching is completely completed to repair the already expanded damage. More importantly, by continuously and dynamically adjusting the passivation gas flow rate to maintain a suppressed state, this process can adapt to the constantly changing sidewall morphology and plasma environment during etching: when the exposed area of ​​the sidewall increases or the corrosivity intensifies, the monitoring parameters will fluctuate accordingly, driving the passivation gas flow rate to automatically increase to strengthen protection; when the sidewall tends to stabilize, the flow rate can be appropriately decreased to avoid excessive passivation affecting etching efficiency. This adaptive adjustment capability is not available in existing fixed-parameter offline passivation schemes. Furthermore, the introduction of a size-locking judgment step means that the termination of etching no longer depends on a fixed time or preset depth, but is based on the actual state of the lateral dimensions of the active region: etching will only stop when the lateral dimensions truly reach the design value and the fluctuations converge to within the allowable range. This fundamentally solves the problem of over-etching or dimensional deviation caused by the inability to perceive the lateral dimensions in real time in traditional processes.

[0023] Based on the above principles, this application can achieve the following beneficial effects, including: First, since the passivation gas is introduced synchronously at the beginning of the transverse drilling and generates a protective layer in situ on the sidewall, rather than being repaired on the surface after the entire etching is completed, it can prevent the quantum well material from being over-consumed from the source and effectively protect the structural integrity of the active region.

[0024] Second, through closed-loop feedback adjustment based on in-situ monitoring parameters, the chemical state of the sidewall is continuously maintained within a stable range that suppresses lateral etching. Regardless of environmental fluctuations or morphological changes during the etching process, the system can adaptively adjust the passivation intensity, thereby achieving dynamic suppression of lateral drilling and avoiding the problems of insufficient or excessive protection that may occur under fixed parameters.

[0025] Third, by acquiring or calculating the current lateral physical size in real time, and precisely terminating the etching process when it reaches the design size and the fluctuations naturally converge, this process improves the lateral size control of the active region from open-loop time prediction to closed-loop state locking, which significantly reduces the size deviation caused by etching rate drift or inter-wafer differences.

[0026] Fourth, the entire etching and passivation process is completed continuously within the same chamber, eliminating the need to remove the wafer for wet processing. This avoids additional risks associated with interface contamination and vacuum loss, and improves process consistency and repeatability.

[0027] In summary, this application fundamentally solves the lateral drilling problem mentioned in the background technology by establishing a complete process chain of sensing, triggering, synchronous intervention, closed-loop maintenance, and size locking, and realizes non-destructive and high-precision etching of the active region of the quantum well of DFB laser.

[0028] In complex plasma etching environments, accurately, sensitively, and reliably identifying the starting point of lateral drilling and avoiding misjudgments or response lags caused by environmental fluctuations or noise interference is fundamental to achieving damage-free etching of the active region of DFB laser quantum wells. Failure to capture early signs of lateral drilling in a timely and accurate manner may lead to accelerated sidewall damage, affecting device performance and yield.

[0029] In response, this application further proposes in-situ monitoring parameters including the characteristic wavelength intensity in the plasma emission spectrum that characterizes the etching byproducts of the quantum well material; and sets the characteristic change that characterizes the start of lateral drilling as follows: the intensity difference of the characteristic wavelength intensity in each of the three consecutive sampling periods exceeds three times the standard deviation of the background noise of the characteristic wavelength intensity.

[0030] Specifically, limiting the in-situ monitoring parameters to the intensity of characteristic wavelengths in the plasma emission spectrum that characterize etching byproducts of the quantum well material means using optical emission spectroscopy (OES) to monitor the plasma within the etching chamber in real time. When the material in the active region of a DFB laser quantum well (e.g., composed of compound semiconductors such as InGaAsP or InAlGaAs) undergoes lateral etching during the etching process, the quantum well material is etched, generating corresponding volatile byproducts. These byproducts emit light of specific wavelengths when excited in the plasma. By monitoring the intensity of these characteristic wavelengths, the etching rate of the quantum well material and the formation of etching products can be directly and in real time reflected. This direct chemical state monitoring method provides highly specific information, effectively distinguishing the lateral etching of the quantum well material from other reactions in the background plasma.

[0031] Building upon this, this application further clarifies the characteristic change that characterizes the start of lateral drilling. This characteristic change is defined as follows: within three consecutive sampling periods, the intensity difference of the characteristic wavelength intensity in each period relative to the previous period exceeds three times the standard deviation of the background noise of the characteristic wavelength intensity. This determination mechanism introduces statistical robustness. First, the introduction of the background noise standard deviation allows the system to quantify and compensate for the inherent random fluctuations and measurement noise of the plasma environment. By setting the threshold to three times the background noise standard deviation, it ensures that only significant and non-random intensity changes are identified as the start of lateral drilling, thus effectively avoiding misjudgments caused by occasional noise or minor fluctuations. Second, the requirement of three consecutive sampling periods further enhances the reliability of the determination. This means that the system will not trigger a response simply because of a single instantaneous intensity jump, but requires that this intensity increase trend have a certain degree of persistence, thereby filtering out transient interference signals and ensuring that only real and continuous lateral drilling behavior is accurately captured.

[0032] Specifically, in the passivation etching process of the lateral sidewalls of the active region of a DFB laser quantum well, to accurately identify the etching byproducts of the quantum well material, this application limits the characteristic wavelengths in the plasma emission spectrum to the 451 nm spectral line corresponding to indium or the 417 nm spectral line corresponding to gallium. Since the active region of a DFB laser quantum well is typically composed of III-V compound semiconductor materials containing indium (In) and gallium (Ga), such as InGaAsP, when these materials are etched in plasma, they release corresponding atoms, which emit light of specific wavelengths under plasma excitation. Choosing 451 nm (indium) and 417 nm (gallium) as characteristic wavelengths can directly and sensitively reflect the etching status of the quantum well active region material, thereby ensuring a high correlation between the monitoring signal and the lateral etching behavior. Besides indium and gallium, other elements constituting the quantum well material, such as phosphorus or arsenic, may also have characteristic spectral lines that can be used for monitoring, but the spectral lines of indium and gallium may exhibit higher sensitivity and specificity under specific etching conditions.

[0033] Meanwhile, this application sets the sampling period for the in-situ monitoring parameters within the range of 100 milliseconds to 500 milliseconds. The sampling period refers to the time interval between the acquisition and processing of characteristic wavelength intensity data by the in-situ monitoring system (e.g., an optical emission spectrometer). Choosing this range of sampling periods aims to balance the ability to capture instantaneous changes in the etching process with signal stability. If the sampling period is too long, early signs of lateral drilling may be delayed in detection, thus affecting the timely intervention of sidewall passivation gas. If the sampling period is too short, insufficient signal integration time may introduce excessive noise, increasing the risk of misjudgment. Therefore, a sampling period of 100 milliseconds to 500 milliseconds ensures that sufficient data quality is maintained for reliable judgment while rapidly responding to changes in the etching state.

[0034] Furthermore, to establish a reliable judgment benchmark, this application specifies that the standard deviation of background noise is calculated from the fluctuation value of the characteristic wavelength intensity within the first 2 seconds after etching starts. The standard deviation of background noise is calculated according to the following formula:

[0035] Where N is the total number of characteristic wavelength intensity data points collected within the first 2 seconds after etching starts, according to a fixed sampling period, and I i Let i be the intensity value of the i-th data point. The arithmetic mean of N data points Background noise standard deviation is a statistical measure of the normal fluctuations in characteristic wavelength intensity when no lateral drilling occurs. By continuously monitoring the characteristic wavelength intensity and calculating the standard deviation of its fluctuations within the first 2 seconds after the start of the etching process—that is, during the stage when lateral drilling has not yet occurred and the etching conditions are relatively stable—a dynamic noise baseline adapted to the current etching environment and equipment status can be obtained. This dynamic baseline is more adaptable than a preset fixed threshold and can effectively distinguish between signal changes caused by normal process fluctuations and significant signal enhancement caused by actual lateral drilling, thus providing an accurate and robust reference for judging the characteristic changes at the start of lateral drilling.

[0036] In the process of introducing sidewall passivation gas to suppress lateral etching, this application further proposes two specific determination methods: a stable state characterizing the stable formation of the sidewall passivation layer and a suppressed state characterizing the effective suppression of lateral etching.

[0037] The stable state characterizing the stable formation of the sidewall passivation layer is defined as follows: the fluctuation amplitude of the in-situ monitoring parameter within five consecutive sampling periods is less than 5% of the average value of the five sampling periods. This technical feature defines a specific quantitative standard for the stable formation of the sidewall passivation layer. During the etching process, after the sidewall passivation gas is introduced, the system needs time to reach dynamic equilibrium, that is, the rate of passivation layer formation and the rate of passivation layer consumption by etching reach equilibrium. By monitoring the fluctuation amplitude of the in-situ monitoring parameter within five consecutive sampling periods and requiring it to be less than 5% of the average value of these five sampling periods, it can be ensured that the passivation layer has stably and uniformly covered the sidewall of the quantum well active region. This judgment method based on statistical stability can effectively avoid misjudgments caused by instantaneous fluctuations or system noise, thereby ensuring the reliability of the passivation effect.

[0038] Meanwhile, the suppression state characterized by effective suppression of lateral etching is defined as follows: the instantaneous value of the in-situ monitoring parameter is consistently lower than the trigger value corresponding to the characteristic change indicating the onset of lateral etching. This technical feature clarifies the criteria for determining effective suppression of lateral etching. During the main etching process, once the characteristic change indicating the onset of lateral etching is detected and a trigger signal is generated, the sidewall passivation gas is introduced. Subsequently, to ensure that lateral etching does not occur again or is effectively suppressed, the in-situ monitoring parameter needs to be continuously monitored. When the instantaneous value of this parameter remains consistently lower than the trigger value corresponding to the initial characteristic change indicating the onset of lateral etching, it indicates that the chemical reaction or physical mechanism leading to lateral etching has been effectively suppressed, and the sidewall is adequately protected. This standard provides a clear, real-time verifiable feedback mechanism, ensuring the anisotropy of the etching process and avoiding lateral damage to the active region of the quantum well.

[0039] During the lateral sidewall passivation etching process of the active region of a DFB laser quantum well, it is necessary to simultaneously introduce sidewall passivation gas into the etching chamber while maintaining a continuous flow of the main etching gas, and to adjust the flow rate of the sidewall passivation gas until the in-situ monitoring parameters reach a stable state characterizing the stable formation of the sidewall passivation layer. However, without a precise and effective flow rate adjustment mechanism, the amount of sidewall passivation gas introduced may not be able to quickly and accurately reach the required level, thus affecting the stable formation efficiency and quality of the sidewall passivation layer, and may even cause instability or over-passivation in the etching process.

[0040] Reference Figure 2 As shown, this application further proposes adjusting the flow rate of the sidewall passivation gas until the in-situ monitoring parameters reach a stable state characterizing the stable formation of the sidewall passivation layer. Specifically, this includes: using a proportional-integral-derivative control method, and calculating the flow rate adjustment of the sidewall passivation gas according to the following formula based on the deviation between the current value of the in-situ monitoring parameters and the target value of the stable state: ΔQ=Kp·e(t)+Ki·∫e(t)dt+Kd·de(t) / dt, where ΔQ is the flow rate adjustment of the sidewall passivation gas, e(t) is the deviation between the current value and the target value of the stable state, Kp, Ki, and Kd are the proportional coefficient, integral coefficient, and derivative coefficient, respectively, and t is time.

[0041] Specifically, Proportional-Integral-Derivative (PID) control is a feedback control algorithm widely used in industrial control. Its core idea is to calculate the control quantity based on the deviation between the current system state and the target state, through the combined action of proportional, integral, and derivative components, to drive the system to approach and stabilize at the target state. In the flow regulation of the sidewall passivation gas, PID control can achieve precise and dynamic adjustment of the flow rate to cope with the complex and variable environment within the etching chamber. During the etching process, in-situ monitoring parameters (such as the characteristic wavelength intensity of etching byproducts in the plasma emission spectrum) can reflect the chemical state of the quantum well sidewall and the formation of the passivation layer in real time. By setting an expected stable state target value and acquiring the current value of the in-situ monitoring parameters in real time, the deviation between the two can be calculated. This deviation is the basis for the PID controller to regulate the flow rate; it quantifies the gap between the current passivation layer formation state and the ideal stable state. The above formula is the mathematical expression of the PID control algorithm. Here, ΔQ represents the flow rate regulation amount of the sidewall passivation gas, that is, the incremental or decremental adjustment of the gas flow rate required at the current moment. e(t) represents the deviation between the current value of the in-situ monitored parameter and the steady-state target value. Kp, Ki, and Kd are the proportional, integral, and derivative coefficients, respectively. These are key parameters in the PID controller and need to be precisely tuned according to the specific etching equipment, process conditions, and material properties to optimize the control effect. The proportional term Kp·e(t) provides the immediate response to the current deviation; the integral term Ki·∫e(t)dt is used to eliminate steady-state errors that may occur during long-term operation, ensuring that the target value can be accurately reached; the derivative term Kd·de(t) / dt is used to predict the trend of deviation changes, thereby adjusting in advance when the deviation changes significantly, effectively suppressing system oscillations, and improving the smoothness and response speed of the control. t represents time, indicating the continuity of the control process.

[0042] In actual etching processes, the etching environment may undergo complex and dynamic changes, making it difficult to accurately control the real-time fluctuation range of in-situ monitoring parameters. If only a single or simple adjustment strategy is adopted, it may not be able to respond to subtle changes in the etching state in a timely and accurate manner, thereby affecting the stability of the sidewall passivation effect and the uniformity of the etching morphology.

[0043] Reference Figure 3As shown, this application also proposes to dynamically adjust the flow rate of the sidewall passivation gas based on the real-time fluctuations of the in-situ monitoring parameters, so as to maintain the in-situ monitoring parameters in a suppressed state that indicates that the lateral etching is effectively suppressed. Specifically, this includes: dividing the real-time fluctuation amplitude of the in-situ monitoring parameters into multiple levels, each level corresponding to a flow rate compensation coefficient for the sidewall passivation gas; when the real-time fluctuation amplitude rises to the upper limit of the current level, the flow rate compensation coefficient is switched to the value corresponding to the next higher level, and the flow rate of the sidewall passivation gas is increased accordingly; when the real-time fluctuation amplitude falls below the lower limit of the current level, the flow rate compensation coefficient is switched to the value corresponding to the next lower level, and the flow rate of the sidewall passivation gas is decreased accordingly.

[0044] Specifically, classifying the real-time fluctuation amplitude of in-situ monitoring parameters into multiple levels refers to quantifying and classifying the real-time variation range or fluctuation degree of in-situ monitoring parameters (e.g., the characteristic wavelength intensity of etching by-products in quantum well materials in plasma emission spectra) based on preset etching process experience or experimental data. Each level represents a different state of the current lateral etching activity or the stability of the sidewall passivation layer; for example, it can be set as slight fluctuation, moderate fluctuation, and severe fluctuation. This classification process enables the system to respond differently to changes in etching state at different degrees. Each level corresponds to a flow rate compensation coefficient for the sidewall passivation gas, meaning that a specific flow rate adjustment factor is preset for each fluctuation level. This compensation coefficient is used to guide the increase or decrease in the flow rate of the sidewall passivation gas, ensuring the accuracy and effectiveness of the flow rate adjustment. For example, when the fluctuation amplitude is large, the corresponding compensation coefficient will be larger to achieve a larger increase in flow rate, thereby quickly suppressing lateral etching.

[0045] When the real-time fluctuation amplitude rises to the upper limit of the current level, the system detects an increasing trend in lateral etching activity. At this point, to more effectively suppress lateral etching, the controller switches the flow compensation coefficient to the preset value corresponding to the next higher level. Subsequently, based on the new compensation coefficient, the flow rate of the sidewall passivation gas is increased according to a predetermined algorithm or ratio. This mechanism ensures that the passivation gas is supplied in a timely and sufficient manner when lateral etching intensifies, thereby enhancing the sidewall passivation effect. Conversely, when the real-time fluctuation amplitude drops below the lower limit of the current level, it indicates that lateral etching activity has been effectively suppressed, or the passivation layer has stabilized. To avoid the negative effects of excessive passivation (such as reduced etching rate, poor sidewall morphology, or increased residue), the system switches the flow compensation coefficient to the value corresponding to the next lower level and decreases the flow rate of the sidewall passivation gas accordingly. This dynamic adjustment mechanism helps optimize gas consumption and maintain the balance of the etching process.

[0046] In some embodiments of this application, during the passivation etching of the lateral sidewalls of the active region of the DFB laser quantum well, it is necessary to calculate the current lateral physical size of the active region to determine whether size locking is complete. (Refer to...) Figure 4 As shown, this application further proposes a method for calculating the current lateral physical size of the active region. Specifically, the reflection spectrum of the top of the ridge waveguide in the active region is measured in real time using an in-situ spectroscopic ellipsometer. Based on the linear relationship between the oscillation period in the reflection spectrum and the lateral size of the active region, the current lateral physical size is calculated according to the following formula: W=λ / (2n·Δθ), where W is the current lateral physical size, λ is the wavelength of the measured light, n is the refractive index of the active region material, and Δθ is the angular interval between two adjacent characteristic peaks in the reflection spectrum.

[0047] Among them, the in-situ spectroscopic ellipsometer is a non-contact, non-destructive optical measurement device that can acquire optical information of the sample surface in real time without interrupting the etching process. It acquires spectral data by emitting a polarized beam of light onto the top of the active region's ridge waveguide and analyzing the changes in the polarization state of the reflected light. Its in-situ nature ensures the real-time nature and accuracy of the measurement results, avoiding the cumbersome process and potential errors introduced by removing the sample from the etching chamber for offline measurements. During the etching process, the in-situ spectroscopic ellipsometer continuously focuses the measurement beam onto the top surface of the ridge waveguide in the forming DFB laser quantum well active region. The reflection spectrum is the result of the interaction between light and the waveguide structure, containing information such as the waveguide's geometric dimensions and material optical constants. By continuously acquiring this spectral data, changes in the waveguide's lateral dimensions can be dynamically tracked.

[0048] For ridge waveguide structures, when a measurement beam illuminates its top, the light undergoes multiple reflections and interferences within the waveguide. These interference effects manifest as periodic oscillations in the reflection spectrum. The lateral dimension (width) of the waveguide directly affects the effective propagation path and modes of light within the waveguide, thus altering the oscillation period of the reflection spectrum. This linear relationship between the oscillation period in the reflection spectrum and the lateral dimension of the active region can typically be established through prior theoretical modeling (e.g., based on mode coupling theory or finite element analysis) or experimental calibration, ensuring the reliability of the conversion from spectral data to physical dimensions.

[0049] The formula W=λ / (2n·Δθ) provides a mathematical model for converting optical measurement results into actual physical dimensions. Here, W represents the current lateral physical dimension of the active region, a target parameter that needs to be acquired in real time; λ is the measurement wavelength used by the in-situ ellipsometer, typically chosen for its good penetration and sensitivity to the active region material; n is the refractive index of the active region material at the measurement wavelength, a value that must be precisely known or simultaneously measured by the ellipsometer; and Δθ represents the angular interval between two adjacent characteristic peaks in the reflection spectrum (e.g., the maximum or minimum reflectance), directly reflecting the oscillation period of the spectrum. By extracting Δθ in real time and combining it with the known λ and n, the current lateral physical dimension W can be accurately calculated.

[0050] In actual etching processes, due to measurement noise, environmental disturbances, or minor fluctuations in the etching process, the lateral physical size of the active region may fluctuate slightly around the design size. This application further proposes a natural convergence interval for the current lateral physical size to enter the range of allowable fluctuations. Specifically, the difference between the maximum and minimum values ​​of the current lateral physical size in the measurements at ten consecutive sampling times is less than 1 nanometer, and the difference between the arithmetic mean of the ten measurements and the design size is less than one percent of the arithmetic mean.

[0051] Specifically, the difference between the maximum and minimum values ​​of the current lateral physical dimension across ten consecutive sampling times is less than 1 nanometer. This condition aims to define the short-term stability of the etching process. Even at extremely low etching rates, slight fluctuations in dimensional measurements may still occur near the end of etching. By setting the range of measurement fluctuations (the difference between the maximum and minimum values) across multiple consecutive sampling times (e.g., ten sampling times) to be less than a very small threshold (e.g., 1 nanometer), transient noise or minor disturbances can be effectively filtered out. This ensures that when dimensional locking is determined to be complete, the lateral dimension of the active region has reached a highly stable state, and there are no longer significant etching or passivation effects causing dimensional changes. This indicates that the etching process has entered a very fine equilibrium stage. Simultaneously, the difference between the arithmetic mean of the ten measurements and the design dimension is less than one-hundredth of the arithmetic mean. This condition ensures the final dimensional accuracy of the active region. After the etching process stabilizes, further confirmation is needed to verify whether the current dimension accurately meets the expected design dimension. By calculating the arithmetic mean of measurements taken at multiple consecutive sampling times (e.g., ten sampling times) and comparing it to a preset design dimension, and requiring the difference between the two to be less than one percent of the arithmetic mean, it can be ensured that the final lateral physical dimension is not only stable but also highly close to the design target, meeting the high dimensional accuracy requirements of DFB lasers. This percentage setting allows for a certain relative error while ensuring that the absolute error is within an acceptable range, balancing accuracy and the feasibility of actual manufacturing processes.

[0052] In the lateral sidewall passivation etching process of the active region of a DFB laser quantum well, the selection and ratio of the main etching gas and the sidewall passivation gas are crucial for achieving precise vertical etching, effectively suppressing lateral drilling, and ensuring the integrity of the active region morphology.

[0053] This application further proposes that the main etching gas comprises a mixture of chlorine and boron trichloride, wherein the volumetric flow rate ratio of chlorine to boron trichloride is 1:2 to 2:1, and this ratio is pre-selected based on the material composition of the quantum well active region and the required anisotropy of the etching morphology. Simultaneously, the sidewall passivation gas comprises methane or hydrogen. When methane is used, its volumetric flow rate is 10% to 30% of the total volumetric flow rate of the main etching gas; when hydrogen is used, its volumetric flow rate is 5% to 15% of the total volumetric flow rate of the main etching gas.

[0054] Specifically, the main etching gas consists of chlorine (Cl2) and boron trichloride (BCl3). Chlorine, as the primary etchant, provides highly efficient chemical etching capabilities, particularly suitable for III-V compound semiconductor materials, contributing to high etching rates and good vertical etching performance. Boron trichloride plays multiple roles in the etching process. It effectively removes the native oxide layer on the material surface, ensuring uniformity and consistency in etching. Simultaneously, its chlorine atoms participate in the etching reaction, further enhancing etching efficiency. By precisely controlling the volumetric flow rate ratio of chlorine to boron trichloride within the range of 1:2 to 2:1, optimization can be achieved based on the specific material composition of the DFB laser quantum well active region (e.g., InP-based or GaAs-based) and the desired etching morphology for the ridge waveguide structure (e.g., the steepness of the sidewalls, the flatness of the bottom). For example, for materials containing more oxides or scenarios requiring stronger chemically assisted etching, the proportion of boron trichloride can be appropriately increased; while for scenarios seeking higher etching rates and anisotropy, the proportion of chlorine can be adjusted. This adjustable gas mixture ratio allows the etching process to flexibly adapt to different material systems and design requirements.

[0055] In addition, methane (CH4) or hydrogen (H2) can be used as the sidewall passivation gas. These passivation gases are introduced simultaneously while the main etching gas is continuously fed in for vertical etching. They decompose in the plasma environment, forming hydrocarbon or hydrogen radicals. These radicals adsorb and deposit on the sidewalls of the quantum well active region, forming a dense polymer or hydride passivation layer. This passivation layer effectively blocks the lateral erosion of the sidewalls by the main etchant, thereby suppressing lateral drilling behavior and ensuring good verticality and morphology of the ridge waveguide structure. When using methane as the sidewall passivation gas, its volumetric flow rate is precisely controlled between 10% and 30% of the total volumetric flow rate of the main etching gas. This range aims to ensure the formation of a passivation layer of sufficient thickness and stability to effectively suppress lateral etching, while avoiding excessively thick passivation layers due to excessive methane flow, which could affect the vertical etching rate or generate unnecessary residues at the etching bottom. If hydrogen is chosen as the sidewall passivation gas, its volumetric flow rate is 5% to 15% of the total volumetric flow rate of the main etching gas. The passivation layer formed by hydrogen is typically thin and easy to remove, yet still provides effective lateral etching suppression. A lower flow rate helps achieve passivation while reducing contamination of the etching chamber and the difficulty of subsequent cleaning.

[0056] In actual etching processes, a single etching gas combination may struggle to effectively control plasma stability while ensuring high etching rates and good anisotropy, and to guarantee uniform and dense deposition of the passivation layer on the sidewalls of the quantum well active region. This could potentially affect the final performance and yield of the DFB laser. This application further proposes a technical solution to optimize the composition of the etching and passivation gases. Specifically, the main etching gas also includes argon, with an argon volumetric flow rate of 5%–20% of the total main etching gas volumetric flow rate. Simultaneously, the sidewall passivation gas includes nitrogen, which is introduced simultaneously with methane or hydrogen, with a nitrogen volumetric flow rate of 20%–40% of the total sidewall passivation gas volumetric flow rate.

[0057] Argon is introduced into the main etching gas. As an inert gas, argon does not directly participate in the chemical reaction during plasma etching, but rather assists in etching through physical bombardment. Argon ions (Ar+) have a large mass, which can effectively bombard the surface of the quantum well active region, enhancing the physical composition of the etching and thus helping to improve the anisotropy of the etching, ensuring the perpendicularity of the ridge waveguide structure. In addition, argon helps to stabilize the plasma discharge, providing a stable ion source and reducing plasma inhomogeneity, thereby improving the repeatability and consistency of the etching process. The volumetric flow rate of argon is 5% to 20% of the total volumetric flow rate of the main etching gas. This proportion is designed to balance the physical bombardment effect with the chemical reaction efficiency of the active etching gas, avoiding excessive dilution that would reduce the etching rate, while ensuring plasma stability.

[0058] Nitrogen gas is introduced into the sidewall passivation gas, simultaneously introduced with methane or hydrogen. Its main function is to optimize the composition and performance of the passivation layer. Nitrogen can react with etching products or sidewall materials to form nitrogen-containing passivation layers, such as nitrides. These nitrogen-containing passivation layers typically exhibit higher chemical stability and lower defect density, significantly improving the compactness and barrier properties of the passivation layer, thereby more effectively suppressing lateral drilling behavior. Furthermore, nitrogen can alter the types and concentrations of free radicals in the plasma, affecting the deposition efficiency of methane or hydrogen on the sidewall and the quality of the passivation layer, forming a composite passivation layer that combines the advantages of different passivating agents, providing more comprehensive sidewall protection. The nitrogen volumetric flow rate is 20%–40% of the total volumetric flow rate of the sidewall passivation gas. This optimized range aims to maximize the improvement effect of nitrogen on the passivation layer performance without interfering with the function of methane or hydrogen as the primary passivating agent.

[0059] To further illustrate the technical effectiveness of the lateral sidewall passivation etching process for the active region of the DFB laser quantum well provided in this application, a comparative experiment was also designed in this embodiment. The experiment used epitaxial wafers from the same batch, with the InGaAsP / InP multi-quantum-well active layer and upper grating layer already grown. The wafers were divided into experimental and control groups, each containing 10 dies with the same structure. The experimental group underwent ridge waveguide etching using the etching process described in this application, while the control group underwent etching with the same structure using a traditional offline passivation process after etching.

[0060] Both sets of experiments used the same etching equipment. The main etching gas was a mixture of chlorine and boron trichloride (volume flow ratio 1:1). The chamber pressure was 5 mTorr. The upper electrode power was 500 W. The lower electrode bias power was 150 W. The etching target depth was 800 nm.

[0061] The experimental group adopted the process described in this application, and the specific steps are as follows: After etching is initiated, the intensity of the 417 nm spectral line corresponding to gallium in plasma emission spectroscopy (OES) is acquired in real time as an in-situ monitoring parameter. The sampling period is 200 milliseconds, and the standard deviation of the background noise is calculated from the fluctuation value of the 417 nm spectral line intensity within the first 2 seconds after etching is initiated.

[0062] When the intensity difference of the 417nm spectral line exceeds three times the standard deviation of the background noise in each of three consecutive sampling periods, lateral drilling is determined to have started, and a trigger signal is generated. Actual measurements show that this trigger signal is generated 8.2 seconds after the main etching begins.

[0063] In response to the trigger signal, while maintaining the continuous flow of the main etching gas, methane (CH4), a sidewall passivation gas, is simultaneously introduced into the etching chamber at an initial flow rate of 20% of the total volumetric flow rate of the main etching gas. A proportional-integral-derivative (PID) control method is employed, dynamically adjusting the methane flow rate based on the deviation between the 417 nm spectral line intensity and the steady-state target value. The steady-state target value is defined as the fluctuation amplitude of the 417 nm spectral line intensity within five consecutive sampling periods being less than 5% of the average of those five periods.

[0064] After the steady state was achieved, it was measured to be achieved 12.5 seconds after the main etching started. Based on the real-time fluctuation of the 417nm spectral line intensity, the methane flow rate was dynamically adjusted using a graded compensation method to keep the spectral line intensity always below the trigger value.

[0065] The current lateral physical dimensions of the active region (i.e., the bottom width of the ridge waveguide) are calculated in real time using an in-situ spectroscopic ellipsometer. When this dimension reaches the design dimension of 2.0 μm, and the difference between the maximum and minimum values ​​among the measurements at ten consecutive sampling times is less than 1 nm, and the difference between the arithmetic mean of these ten measurements and the design dimension is less than 1% of the arithmetic mean, the dimension is considered to be locked, all gas flow is stopped, and etching is terminated.

[0066] The average total etching time for the 10 dies in the experimental group was 47.3 seconds.

[0067] The control group used a traditional etching followed by offline passivation process: The same etching equipment and main etching gas parameters as the experimental group were used for etching. The etching time was fixed at 47.3 seconds (consistent with the average time of the experimental group). No in-situ monitoring was performed during the etching process, and no passivation gas was introduced.

[0068] After etching is completed, the wafer is removed from the etching chamber and the etched surface is soaked in sulfuric acid and hydrogen peroxide solution for 60 seconds to remove oxides. Then, the etched surface is soaked in ammonium polysulfide solution for 120 seconds to form a sulfide protective layer on the etched surface.

[0069] After rinsing with deionized water, the material is dried with nitrogen to complete the passivation process.

[0070] After etching, the following key indicators of the two sets of samples were tested: Lateral erosion depth: The ridge waveguide cross-section of each die was photographed and measured using a scanning electron microscope (SEM), and the maximum lateral erosion depth inward from the active region sidewall was calculated. The average value of 10 dies in each group was taken.

[0071] The lateral dimension deviation of the active region: Based on the design dimension of 2.0μm, the absolute value of the absolute deviation between the actual bottom width of the ridge waveguide and the design dimension is measured, and the average value of 10 dies in each group is taken.

[0072] Sidewall roughness: The root mean square roughness (Rq) of the sidewall of the ridge waveguide was measured using atomic force microscopy (AFM). The scanning range was 1 μm × 1 μm, and the average value of 10 dies was taken for each group.

[0073] Photoluminescence intensity: The peak intensity of the active region was measured using a micro-area photoluminescence testing system (excitation wavelength 532nm, power 10mW). The relative intensity retention rate was calculated with the photoluminescence intensity of the unetched original epitaxial wafer as 100%.

[0074] The test results are shown in Table 1: Table 1 Lateral drilling depth (nm) 2.3 18.7 Reduced by 87.7% Lateral size deviation of active region (nm) ±1.2 ±5.6 Shrink by 78.6% Sidewall roughness Rq (nm) 0.8 1.9 Reduced by 57.9% Relative intensity retention rate of photoluminescence (%) 96.4% 78.2% Increased by 23.3% The comparative experimental results above show that the technical solution of this application has produced unexpected technical effects compared with the prior art, specifically in the following aspects: First, regarding lateral drilling suppression, the experimental group's lateral drilling depth was only 2.3 nm, while the control group's was as high as 18.7 nm, a difference of more than eight times. This indicates that this application, by simultaneously introducing passivation gas during the etching process and performing closed-loop feedback adjustment based on in-situ monitoring parameters, can form an effective in-situ protective layer as soon as the quantum well sidewalls are exposed, thereby preventing the lateral diffusion of etching gas into the sidewall interior from the source. In contrast, the offline passivation scheme of the control group, although capable of chemically repairing the etched surface after etching, cannot undo the lateral material loss that has already occurred during etching. Its drilling depth is close to 2% of the total width of the active region, which is sufficient to significantly affect luminescence efficiency and mode characteristics.

[0075] Second, regarding dimensional control accuracy, the lateral dimensional deviation of the active region in the experimental group was controlled within ±1.2 nm, while the deviation in the control group was ±5.6 nm. The fundamental reason for this difference lies in the introduction of a dimensional locking judgment step in this application. Etching is terminated only when the lateral physical dimension reaches the design dimension and enters the natural convergence range, rather than relying on a fixed etching time. Traditional processes, unable to perceive the true state of the lateral dimension in real time, cannot eliminate dimensional uncertainties caused by etching rate drift and inter-wafer differences, even with precise etching time control. This application transforms dimensional control from open-loop time prediction to closed-loop state locking, achieving sub-nanometer dimensional accuracy.

[0076] Third, regarding sidewall quality, the sidewall roughness of the experimental group (0.8 nm) was significantly lower than that of the control group (1.9 nm). This is because the passivation gas introduced simultaneously in this application not only inhibited lateral drilling but also formed a uniform passivation protective layer on the sidewall surface. This protective layer continues to function during subsequent etching, preventing the sidewall from becoming roughened due to repeated exposure to plasma bombardment. Although the control group generated a sulfide protective layer through solution immersion in post-processing, this protective layer was formed on the already roughened sidewall surface and could not smooth out the original roughness.

[0077] Fourth, regarding the photoluminescence intensity retention rate, which best reflects the degree of damage to the active region, the experimental group achieved 96.4%, meaning that the etching process caused almost no substantial damage to the luminescence efficiency of the active region; while the control group only achieved 78.2%, indicating that as much as 21.8% of the luminescence efficiency was lost during the etching process due to lateral drilling and sidewall damage. This difference directly proves the technical effectiveness of the non-destructive etching method of this application, while traditional offline passivation methods cannot achieve true non-destructive etching.

[0078] Obviously, the above embodiments are merely illustrative examples for clear explanation and are not intended to limit the implementation. Those skilled in the art will recognize that other variations or modifications can be made based on the above description. It is neither necessary nor possible to exhaustively list all possible implementations here. However, obvious variations or modifications derived therefrom are still within the scope of protection of this invention.

Claims

1. A passivation etching process for the lateral sidewalls of the active region of a DFB laser quantum well, characterized in that, Includes the following steps: The main etching gas is introduced into the etching chamber to perform vertical main etching on the active region of the quantum well of the distributed feedback laser to form a ridge waveguide structure. During the main etching process, in-situ monitoring parameters characterizing the chemical state of the quantum well sidewalls are acquired in real time. When the in-situ monitoring parameters show a characteristic change indicating the start of lateral drilling, lateral drilling behavior is determined to have occurred on the sidewall, and a trigger signal is generated. The in-situ monitoring parameters include the characteristic wavelength intensity in the plasma emission spectrum, which characterizes the etching byproducts of the quantum well material. The characteristic change indicating the start of lateral drilling is defined as follows: the intensity difference of the characteristic wavelength intensity in each of the three consecutive sampling periods exceeds three times the standard deviation of the background noise of the characteristic wavelength intensity. The characteristic wavelength includes the 451 nm spectral line corresponding to indium or the 417 nm spectral line corresponding to gallium. The sampling period is 100 milliseconds to 500 milliseconds. The standard deviation of the background noise is calculated from the fluctuation value of the characteristic wavelength intensity within the first two seconds after etching begins. In response to the trigger signal, while maintaining the continuous flow of the main etching gas, the sidewall passivation gas is simultaneously introduced into the etching chamber, and the flow rate of the sidewall passivation gas is adjusted until the in-situ monitoring parameters reach a stable state that characterizes the stable formation of the sidewall passivation layer. The stable state that characterizes the stable formation of the sidewall passivation layer is: the fluctuation amplitude of the in-situ monitoring parameters within five consecutive sampling periods is less than five percent of the average value of the five sampling periods. Under stable conditions, in-situ monitoring parameters are continuously acquired in real time, and the flow rate of the sidewall passivation gas is dynamically adjusted based on the real-time fluctuations of the in-situ monitoring parameters to maintain the in-situ monitoring parameters in a suppressed state that represents the effective suppression of lateral etching. The suppressed state that represents the effective suppression of lateral etching is: the instantaneous value of the in-situ monitoring parameter is always lower than the trigger value corresponding to the characteristic change that represents the start of lateral drilling. Calculate the current lateral physical size of the active region; when the current lateral physical size reaches the design size representing the completion of the active region design, and the current lateral physical size enters the natural convergence interval representing the allowable fluctuation range, it is determined that the size locking is complete; the current lateral physical size enters the natural convergence interval representing the allowable fluctuation range, specifically: in the measurement values ​​of the current lateral physical size at ten consecutive sampling times, the difference between the maximum and minimum values ​​is less than 1 nanometer, and the difference between the arithmetic mean of the ten measurement values ​​and the design size is less than one-hundredth of the arithmetic mean; In response to the determination that size locking is complete, the main etching gas and sidewall passivation gas are stopped, and the damage-free etching of the active region of the DFB laser quantum well is completed.

2. The passivation etching process for the lateral sidewalls of the active region of the DFB laser quantum well according to claim 1, characterized in that: Adjusting the flow rate of the sidewall passivation gas until the in-situ monitoring parameters reach a stable state characterizing the stable formation of the sidewall passivation layer involves: using a proportional-integral-derivative control method, and calculating the flow rate adjustment of the sidewall passivation gas according to the following formula based on the deviation between the current value of the in-situ monitoring parameters and the target value of the stable state: ΔQ=Kp·e(t)+Ki·∫e(t)dt+Kd·de(t) / dt, where ΔQ is the flow rate adjustment of the sidewall passivation gas, e(t) is the deviation between the current value and the target value of the stable state, Kp, Ki, and Kd are the proportional coefficient, integral coefficient, and derivative coefficient, respectively, and t is time.

3. The passivation etching process for the lateral sidewalls of the active region of the DFB laser quantum well according to claim 1, characterized in that: Based on the real-time fluctuations of in-situ monitoring parameters, the flow rate of the sidewall passivation gas is dynamically adjusted to maintain the in-situ monitoring parameters in a suppressed state that indicates effective suppression of lateral etching. Specifically, this includes: dividing the real-time fluctuation amplitude of the in-situ monitoring parameters into multiple levels, with each level corresponding to a flow rate compensation coefficient for the sidewall passivation gas; when the real-time fluctuation amplitude rises to the upper limit of the current level, the flow rate compensation coefficient is switched to the value corresponding to the next higher level, and the flow rate of the sidewall passivation gas is increased accordingly; when the real-time fluctuation amplitude falls below the lower limit of the current level, the flow rate compensation coefficient is switched to the value corresponding to the next lower level, and the flow rate of the sidewall passivation gas is decreased accordingly.

4. The passivation etching process for the lateral sidewalls of the active region of the DFB laser quantum well according to claim 1, characterized in that: The calculation of the current lateral physical size of the active region specifically includes: measuring the reflection spectrum at the top of the ridge waveguide in the active region in real time using an in-situ spectroscopic ellipsometer; and calculating the current lateral physical size according to the linear relationship between the oscillation period in the reflection spectrum and the lateral size of the active region using the following formula: W=λ / (2n·Δθ), where W is the current lateral physical size, λ is the wavelength of the measured light, n is the refractive index of the active region material, and Δθ is the angular interval between two adjacent characteristic peaks in the reflection spectrum.

5. The passivation etching process for the lateral sidewalls of the active region of the DFB laser quantum well according to claim 1, characterized in that: The main etching gas includes a mixture of chlorine and boron trichloride, wherein the volume flow rate ratio of chlorine to boron trichloride is 1:2 to 2:1, and the ratio is pre-selected according to the material composition of the active region of the quantum well and the anisotropy of the required etching morphology. The sidewall passivation gas includes methane or hydrogen. When methane is used, its volume flow rate is 10% to 30% of the total volume flow rate of the main etching gas. When hydrogen is used, its volume flow rate is 5% to 15% of the total volume flow rate of the main etching gas.

6. The passivation etching process for the lateral sidewalls of the active region of the DFB laser quantum well according to claim 5, characterized in that: The main etching gas also includes argon, with a volumetric flow rate of 5% to 20% of the total volumetric flow rate of the main etching gas; the sidewall passivation gas also includes nitrogen, which is introduced simultaneously with methane or hydrogen, with a volumetric flow rate of 20% to 40% of the total volumetric flow rate of the sidewall passivation gas.

Images