A method for detecting the strength of parasitic feedback of a butt joint end face of a butt joint structure DFB laser chip

Abstract

The application provides a method for detecting parasitic feedback intensity of a butt joint end face of a butt joint structure DFB laser chip, belongs to the technical field of semiconductor optoelectronic device manufacturing, and is used for solving the technical problem that the existing DFB laser chip provided with an end face butt joint structure generates parasitic light feedback at the butt joint end face, and influences spectral characteristics and modulation bandwidth; the application adopts a spectrometer to perform high-precision spectral testing on the DFB laser chip to obtain sampling data, analyzes and calculates the jitter state formed by the high and low light intensity of different wavelength points of the obtained test spectrum appearance, judges whether internal end face light feedback exists in the butt joint structure, analyzes the relative intensity of the end face feedback of the butt joint structure, and obtains the parasitic feedback intensity. The application can judge whether internal end face light feedback exists in the butt joint structure and analyze the relative intensity of the butt joint end face feedback, thereby providing an optimization direction and basis for end face butt joint process and optical waveguide design; and the application does not need to increase new testing equipment, and has good universality.

Description

Technical Field

[0001] This invention relates to the technical field of semiconductor optoelectronic device manufacturing, and in particular to a method for detecting the parasitic feedback intensity of the docking end face of a DFB laser chip with a docking structure. Background Technology

[0002] Distributed feedback (DFB) semiconductor lasers, due to their high modulation bandwidth, high output power, and low cost, have important applications in optical communication, spectral analysis, and lidar. Their core technology involves epitaxy of materials using bandgap engineering theory, achieving single-mode lasing while suppressing noisy sidemodes through an embedded grating, fabricating an optical resonator with parameters meeting the relevant application targets using semiconductor processes on an epitaxial wafer, then cleaving the wafer to form a single laser chip, and finally using optical coatings to control the reflectivity at both ends of the chip and protect the exposed epitaxial material surfaces. Figure 1 The flowchart shows the fabrication process of a DFB laser, including primary epitaxy, grating fabrication, secondary epitaxy, semiconductor fabrication, cleaving, and coating. The semiconductor process enables the design of DFB laser materials and cavity structure, ultimately forming a laser chip that meets the requirements of optoelectronic parameters. Figure 2 The diagram shows the structure of a DFB laser after coating. The central part has an active light-emitting region, and both the backlight and light-emitting surfaces at both ends are coated. Holes injected from the P-side and electrons injected from the N-side recombine in the active region. The energy generated by the recombination is released in the form of light waves, which are then lased in the optical resonant cavity formed by the coated chips at both ends, and finally emitted as laser light at the light-emitting end face in the form of a specific wavelength.

[0003] In traditional DFB laser applications, InGaAsP (indium gallium arsenide phosphide) is usually chosen as the quantum well material for the active light-emitting region to improve chip reliability. This is because InGaAsP is not easily oxidized, and its material purity can be maintained during material epitaxy and fabrication. DFB lasers made of this material have a low probability of being oxidized during use after the optical cavity end face is protected with a coating. Therefore, the related optical chips have high reliability and are less prone to burn-out failure.

[0004] However, with the continuous improvement of computing power for large AI models and the increasing bandwidth requirements of 5G communication technology, the shortcomings of the InGaAsP material system have gradually become apparent. The conduction band barrier of this material's quantum well has a small offset, while the valence band barrier has a large offset. This makes it easy for electrons to overflow when the optical chip operates at higher temperatures, leading to a decrease in chip slope efficiency. It also reduces the migration rate of holes with larger effective mass, thus preventing an increase in modulation bandwidth, severely limiting its application in high-speed and 5G communications. In contrast, the band structure of AlGaInAs (aluminum gallium indium arsenide) quantum wells is more conducive to chip applications in high-temperature, high-bandwidth environments. However, the Al material in AlGaInAs is easily oxidized, resulting in significant reliability risks even with optical coatings protecting the light-emitting end face. Figure 3 This is a comparison chart of the quantum well band shifts for two material systems. (Through...) Figure 3 It can be seen that the shallower conduction band potential well of AlGaInAs is more conducive to the propagation of holes with large effective mass in the well, which is beneficial to high-frequency modulation. Its valence band potential well depth is deeper than that of InGaAsP, which is more conducive to reducing the overflow loss of electrons with small effective mass under high temperature conditions, and is beneficial to improving the high temperature characteristics of the chip.

[0005] The method of secondary epitaxy at both ends of an AlGaInAs laser with InGaAsP protective end faces can greatly improve the reliability of the chip without affecting the high temperature and high bandwidth characteristics of the AlGaInAs active region material system. However, this method will inevitably introduce optical feedback at the end face. The modulation bandwidth of high-speed optical chips is extremely sensitive to this kind of feedback. Excessive feedback energy will reduce the modulation bandwidth, a key parameter of high-speed DFB laser chips, and cause abnormal fluctuations in the output spectrum, or even mode jumps. There is no effective solution in the existing testing methods to determine the existence of this feedback and how to characterize its intensity. This makes it impossible to effectively detect and control, which seriously affects the device performance and production yield. Summary of the Invention

[0006] To address the technical problem of parasitic optical feedback at the docking face of existing DFB laser chips with end-face mating structures, which occurs due to differences in the effective refractive index of the optical wave path caused by variations in the materials and waveguide structure on both sides of the docking face, thus affecting the spectral characteristics and modulation bandwidth of the chip, this invention proposes a method for detecting the intensity of parasitic feedback at the docking face of a DFB laser chip with a docking structure. This method uses spectral testing and specific spectral analysis to determine whether feedback exists at the internal interface and to qualitatively assess the intensity of this feedback. This invention also provides an application of the method for measuring and analyzing the feedback intensity at the docking face of a DFB laser with a docking structure.

[0007] To achieve the above objectives, the technical solution of this invention is as follows: A method for detecting the parasitic feedback intensity of the docking end face of a DFB laser chip in a docking structure. This method employs a spectrometer to perform high-precision spectral testing on the DFB laser chip to obtain sampling data. The obtained test spectral morphology is analyzed and calculated to determine the jitter state caused by varying point light intensities at different wavelengths. This determines whether internal end-face optical feedback exists in the docking structure, and analyzes the relative intensity of the end-face feedback to obtain the parasitic feedback intensity. This method can determine whether internal end-face optical feedback exists in the docking structure and achieves the purpose of analyzing the relative intensity of end-face feedback, thereby providing optimization directions and basis for end-face docking processes and optical waveguide design.

[0008] A method for detecting the parasitic feedback intensity at the docking end face of a DFB laser chip in a docking structure includes the following steps:

[0009] Step S1: A DFB laser chip with docking structures on the optical end face and the back light end face is fabricated using active light-emitting region epitaxy and docking epitaxy;

[0010] Step S2: The wafer of the prepared DFB laser chip is cleaved and sliced ​​to form a single laser chip, and the single laser chip is placed on the COC carrier by eutectic bonding and the positive and negative electrodes of the COC carrier are connected by gold wires.

[0011] Step S3: Fix the COC carrier on the test stage, attach the positive and negative electrodes of the power probe to the corresponding positive and negative electrodes of the COC carrier, and simultaneously couple the light emitted by the single laser chip into the spectrometer using a wedge-shaped end-face coupling optical fiber at the light-emitting end face of the single laser chip.

[0012] Step S4: Set the spectrometer test parameters, stabilize the COC carrier with DC power, and read the spectrum of a single laser chip;

[0013] Step S5: Analyze and calculate the read spectrum to determine whether there is parasitic optical feedback at the docking end face; compare the Ripple intensity of the optical feedback spectrum of single laser chips from different design batches to obtain the comparison relationship of the parasitic intensity of the docking end face feedback in different designs.

[0014] Preferably, the analysis of the test spectrum includes: 1) analyzing the spectral jitter on the right side of the lasing wavelength due to feedback from the light-emitting end face. If a new spectral jitter is superimposed on the normal light-emitting end face feedback jitter Ripple ripple, it is determined that the DFB laser chip has end face parasitic optical feedback of the docking structure; 2) by reading the intensity of the parasitic jitter of the Ripple ripple caused by the parasitic optical feedback, i.e., the height difference h between the peak and trough of the superimposed waveform, and comparing it with the different intensity differences h obtained from different batches of chips, the high-low comparison relationship of the end face feedback parasitic intensity of different designs is finally obtained.

[0015] Preferably, both the active light-emitting region epitaxy and the docking epitaxy are completed by metal-organic chemical vapor deposition; the active light-emitting region epitaxy adopts an AlGaInAs quantum well structure, and the docking structure adopts an InGaAsP quantum well structure. This can greatly improve the reliability of the chip, while not affecting the high-temperature and high-bandwidth characteristics of the AlGaInAs active region material system.

[0016] Preferably, the wedge-shaped end face coupling fiber has an end face bevel angle of 8°±3°, which can reduce the influence of the fiber's feedback light on the chip's own spectrum without causing excessive energy loss.

[0017] Preferably, the spectrometer test parameters include: spectral scanning sensitivity ≥ 20 seconds / 100 nm; spectral scanning accuracy ≥ 0.01 nm; and total spectral scanning width of lasing wavelength ± 20 nm to lasing wavelength ± 80 nm.

[0018] Preferably, docking structures are provided at both ends of the active light-emitting region. The outer backlight surface of the docking structure on the backlight end face is provided with a high-reflection coating, and the outer side of the docking structure on the emission end face is provided with a low-reflection coating. The effective length of the optical resonant cavity formed by the active light-emitting region and the docking structure at the input end is L1, and the physical length of the DFB laser chip formed by the active light-emitting region and the two docking structures is L2.

[0019] When light waves oscillate within a resonant cavity, they form standing waves. Only wavelengths that meet the resonance condition can exist stably within the optical waveguide and be output through the output end face of the DFB laser chip to form effective optical power. All light waves that can form stable resonance are selected by the grating to form the main lasing peak, while light waves that are not selected by the grating are emitted in the form of small gain peaks. The height of the small gain peaks formed by the feedback from the output end face is called the ripple.

[0020] Preferably, the longitudinal mode spacing d of the gain peak is... for: Where λ is the wavelength of the calculated interval frequency band, ñ is the effective refractive index of the optical waveguide, and L is the effective cavity length of the waveguide resonant cavity;

[0021] The effective refractive index ñ is determined by the lasing wavelength λ of the grating lasing peak of the DFB laser chip. g =2ñ∧ is calculated, where ∧ is the grating period.

[0022] Preferably, the longitudinal mode spacing d of the small peaks of spectral gain is obtained based on the read spectrum. The effective cavity length L is calculated using the effective refractive index ñ. By comparing the effective cavity length L with the lengths of different chip structures, it can be determined which feedback end face is responsible for the gain peak. Furthermore, the grating period can be adjusted to make the grating lasing peak avoid the strongest and most obvious gain center, thus avoiding the impact on the readout accuracy of the Ripple ripple of the gain peak.

[0023] The height of the gain peak is determined by the residual reflectivity of the low-reflectivity coating on the exit end face;

[0024] The high-reflectivity coating on the backlight side of a DFB laser chip typically has a reflectivity greater than 90%. The interface reflectivity of the docking structure on the backlight side is in the low-reflectivity range. Therefore, the docking structure near the backlight side will not have a significant impact on the spectrum. However, the low-reflectivity coating on the light-emitting side is of the same order of magnitude as the docking structure near it. The feedback from the docking structure near the light-emitting side will affect the spectral morphology.

[0025] Preferably, if there is no feedback at the cross-section of the docking structure, the output spectrum will only contain the grating lasing peak and a regular parabolic resonant spectrum. If there is parasitic small feedback at the end face of the docking structure, the optical resonant cavity, the active light-emitting region, and the resonant cavity composed of the two docking structures will each generate a resonant spectrum. Due to dual-spectral interference, in addition to the grating lasing peak, the output spectrum will also contain a resonant spectrum with high and low fluctuations. Moreover, the intensity of the grating lasing peak is much greater than that of the two resonant spectra. According to the principle of two-peak interference, interference fringes will form between the two resonant spectra. The spectral amplitude intensity A after the superposition of the two spectra is: A1 A2 ];in, The amplitude intensity of the resonant spectrum of the resonant cavity including the backlight end-face contact structure. The amplitude intensity of the resonant spectrum of the resonant cavity, which includes the interface structure of the backlight end face and the light-emitting end face. The initial phase of a resonant spectrum, This is the initial phase of another resonant spectrum;

[0026] After determining the reflectivity of the end-face coating of the docking structure, the height of the ripple of the gain peak is determined; the feedback intensity of the docking structure end face is determined based on the difference h between the highest and lowest points of the interference peak in the measured spectrum; amplitude intensity. The resonance spectrum and amplitude intensity The peaks and troughs of the two peaks in the resonance spectrum cancel each other out or overlap and increase. The greater the feedback intensity, the greater the difference h between the peaks and troughs, thus obtaining the parasitic feedback intensity of the docking end face.

[0027] The beneficial effects of this invention are: it can detect whether redundant parasitic feedback light is introduced in the design and process of DFB laser chips. The detectable parameters include: whether there is parasitic feedback in the chip; the relative intensity of multiple batches of parasitic feedback. It can realize the monitoring technology and application of the feedback intensity of the docking surface of the end face secondary epitaxial docking structure DFB laser and the lasing spectrum morphology of the laser.

[0028] This invention utilizes spectral sampling of DFB laser chips, followed by specialized analysis and calculation of the spectral morphology, to determine the presence of internal end-face optical feedback in the mating structure and to analyze the relative intensity of this feedback. This provides optimization directions and a basis for end-face mating processes and optical waveguide design. This invention does not require additional testing equipment; existing general-purpose spectrometers can be used to complete the tests, enabling low-cost and rapid acquisition of test results, and demonstrating good versatility.

[0029] This invention can accurately determine whether there is parasitic optical feedback at the docking end face inside the chip: by analyzing the spectral appearance morphology measured by a spectrometer, it is possible to determine whether the docking process (including wet etching end face morphology and docking epitaxial morphology) has generated harmful parasitic optical feedback.

[0030] Qualitative assessment of the level of parasitic end-face feedback: By analyzing the superimposed waveform and calculating the height difference h between the peaks and troughs of the superimposed waveform, the strength of parasitic feedback can be obtained, allowing for improvements to related processes and designs, and guiding whether the improvement effect is conducive to reducing parasitic end-face feedback.

[0031] Good versatility: This invention can be applied to the monitoring and measurement of parasitic feedback in optical chips that require docking structures, such as high-power DFB (Distributed Feedback) optical chips with docking structure CW (Continuous Wave) and high-speed DML (Directly Modulated Laser) optical chips with docking structure. Attached Figure Description

[0032] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0033] Figure 1 This is a flowchart of the fabrication process for a DFB laser.

[0034] Figure 2 This is a structural diagram of a DFB laser.

[0035] Figure 3 This is a simplified diagram of the quantum well band structure of InGaAsP and AlGaInAs material systems.

[0036] Figure 4 This is a schematic diagram of an end-to-end structured light chip.

[0037] Figure 5 This is a schematic diagram showing the two sets of spectra before and after superposition.

[0038] Figure 6 The simulated spectrum of a structured light chip with parasitic feedback end face.

[0039] Figure 7 A schematic diagram of the end face of the docking surface with different reflection angles.

[0040] Figure 8 This is a schematic diagram of the three-stage extension of the present invention, wherein (a) is the original view and (b) is an enlarged view. Detailed Implementation

[0041] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0042] Example 1

[0043] A method for detecting the intensity of parasitic feedback at the docking end face of a DFB laser chip in a docking structure involves using a spectrometer to perform high-precision spectral testing on the DFB laser chip to obtain sampling data. Then, the obtained test spectral morphology is analyzed and calculated to determine the jitter state caused by varying point light intensities at different wavelengths. This allows for the assessment of whether internal end-face optical feedback exists in the docking structure and the analysis of the relative intensity of the end-face feedback. The testing, calculation, and analysis include the following steps:

[0044] Step S1: A DFB laser chip with docking structures on the optical end face and the back light end face is fabricated by active light-emitting region epitaxy and docking epitaxy.

[0045] Both active light-emitting region epitaxy and docking epitaxy are completed using MOCVD (Metal-Organic Chemical Vapor Deposition) equipment. Depending on the function of the different chips being fabricated, the flow rate and temperature of the relevant material sources are set to obtain semiconductor material structures with different structures. This can achieve high compositional precision and high-quality semiconductor materials, and reduce semiconductor material defects.

[0046] Figure 4 This is a schematic diagram of a laser with a docking structure at both ends. Figure 2 Compared to DFB lasers, this laser has docking structures at both ends of the active light-emitting region. The outer back surface of the docking structure at the input end has a high-reflection coating, while the light-emitting surface at the laser output has a low-reflection coating. Here, L1 is the effective length of the optical resonant cavity, and L2 is the physical length of the optical chip. For example... Figure 4 As shown, the active light-emitting region epitaxially adopts an AlGaInAs quantum well structure, and the docking structure adopts an InGaAsP quantum well structure.

[0047] Step S2: The prepared DFB laser chip wafer is cleaved and sliced ​​to form a single laser chip, and the single laser chip is placed on the COC carrier through a eutectic patch and the positive and negative electrodes of the COC carrier are connected with gold wires.

[0048] The role of the COC carrier is to support the fragile DFB laser chip, increase heat dissipation, facilitate subsequent testing, and at the same time, remove some thermal effects to improve testing accuracy.

[0049] Step S3: Fix the COC carrier on the test stage, attach the positive and negative electrodes of the power probe to the corresponding positive and negative electrodes of the COC carrier, and simultaneously couple the light emitted by the single laser chip into the spectrometer using a wedge-shaped end-face coupled optical fiber at the light-emitting end face of the single laser chip.

[0050] The test stage can fix the COC carrier, and at the same time, the test stage can accurately control the temperature of the entire COC carrier to improve the test accuracy.

[0051] The wedge-shaped end face coupling fiber has an end face bevel angle of 8°±3°. This angle can reduce the influence of the fiber's feedback light on the chip's own spectrum without causing excessive energy loss.

[0052] Step S4: Set the spectrometer test parameters, stabilize the COC carrier with DC power, and read the spectrum of a single laser chip.

[0053] The spectrometer settings are as follows: spectral scanning sensitivity ≥ 20 seconds / 100nm; spectral scanning accuracy ≥ 0.01nm; total spectral scanning width is lasing wavelength ±20nm to lasing wavelength ±80nm. This spectral scanning setting can meet the sampling accuracy required for spectral analysis and improve the accuracy of the analysis.

[0054] Step S5: Analyze and calculate the read spectrum to determine whether there is parasitic optical feedback at the docking end; further analyze the ripple levels of the optical feedback spectrum of individual laser chips from different design batches to provide data guidance and support for chip design.

[0055] When light waves oscillate within a resonant cavity, they form standing waves. Only wavelengths that meet the resonance condition will remain stable within the optical waveguide. Wavelengths that do not meet the resonance condition will disappear due to absorption loss, scattering, and other factors. Wavelengths that meet the resonance condition can ultimately be output through the output facet of the DFB laser chip, forming effective optical power. All light waves that can form stable resonance are then selected by a grating to form the main lasing peak. Light waves not selected by the grating are emitted as gain peaks. The height of these gain peaks is determined by the residual reflectivity of the antireflection coating on the output facet (AR facet). Generally, the reflectivity requirement for DFB laser chip applications is less than 0.01%. The height of the gain peaks formed by this facet feedback is generally called the ripple. A smaller ripple indicates a lower residual reflectivity on the AR facet, which is more beneficial to the stability and efficiency of the DFB laser chip. Figure 5 This is an explanation of the spectrum of the optical chip.

[0056] The lasing wavelength λ of the grating lasing peak g As shown below:

[0057] λ g =2ñ∧ ; (1)

[0058] Where, λ g λ is the lasing wavelength of the DFB laser chip, in nm; ñ is the effective refractive index of the optical waveguide; and ∧ is the grating period. g It can be directly read from the obtained spectrum, where ∧ is the design value, and the effective refractive index ñ can be directly calculated using the above formula.

[0059] Longitudinal mode spacing d of gain peak As shown below:

[0060] ; (2)

[0061] Where λ is the wavelength of the calculated interval frequency band in nm, ñ is the effective refractive index of the optical waveguide, and L is the effective optical path length of the waveguide resonant cavity, i.e., the cavity length. (like Figure 5 The values ​​shown in the figure), λ (the wavelength of interest) can be obtained from the tested spectrum, and the effective refractive index ñ is calculated by formula (1).

[0062] As can be seen from formula (1), the lasing wavelength λ of the DFB laser chip is determined by setting the grating period ∧. g Meanwhile, the effective refractive index ñ of the chip is obtained, and the longitudinal mode spacing d of the small peak of spectral gain can be read out according to formula (2). The effective cavity length L is calculated using the effective refractive index ñ calculated in formula (1). By comparing the effective cavity length L with the lengths of different chip structures, it can be proved which feedback end face is responsible for the gain peak. Furthermore, the grating period can be adjusted to make the grating lasing peak avoid the strongest and most obvious gain center, thus avoiding the influence on the Ripple readout accuracy of the gain peak.

[0063] The backlight end face (HR end) of the DFB laser chip is coated with a high-reflectivity film, which generally has a reflectivity greater than 90%. The interface reflectivity of the docking structure at the backlight end is in the low-reflectivity range. Therefore, the docking interface 1 near the HR end will not have a significant impact on the spectrum. The low-reflectivity film layer of the AR coating on the light-emitting end face is of the same order of magnitude as the docking end face 2 nearby. Therefore, the feedback of this end face (docking end face 2) will affect the spectral morphology. Since the docking structures at both ends use the same process and design, analyzing the feedback intensity of one end face alone will not affect the result of judging the feedback intensity of both docking end faces at the same time.

[0064] If there is no feedback from the docking end face, the output spectrum will only contain the grating lasing peak and Figure 8 In Figure a, the regular parabolic resonance spectrum type 1, if there is a small parasitic feedback at the docking end face, then resonant cavities L1 and L2 will each generate a set of resonance spectra. and Therefore, due to Figure 5 The dual-spectral interference shown will produce an output spectrum that includes not only the grating lasing peak but also... Figure 8 In Figure a, spectral type 3 or type 2 provides a clearer morphology, similar to a simulated spectrum. Figure 6 The high and low vibrations, and from Figure 8 It can be seen that the intensity of the grating lasing peak is much greater than the resonance spectrum produced by these two resonant cavities. and resonance spectrum Therefore, the resonance spectrum and It will not have a significant impact on the grating lasing peak, but the resonance spectrum and resonance spectrum According to the principle of two-peak interference, interference fringes will be formed between the two peaks. The spectral amplitude intensity A after the two spectra are superimposed is shown below:

[0065] A1 A2 (3)

[0066] in, The amplitude intensity of the resonant spectrum of the resonant cavity including the backlight end-face contact structure. Let A be the amplitude intensity of the resonant spectrum of the resonant cavity containing the backlight end face and the light-emitting end face contact structure. The spectral amplitude intensity A is... and The intensity of the spectral amplitude after coherent superposition, Let be the initial phase of the resonant spectrum 'a'. Let be the initial phase of the resonant spectrum b. This formula is only intended to illustrate the coherence characteristics of the two spectra.

[0067] The amplitude intensity can be obtained from formula (3). and The final spectrum after superposition should exhibit a cosine-wave pattern in intensity. Figure 5 The image shows a schematic diagram of the superimposed spectra of two sets of spectra with similar intensities. Figure 5 It can be seen that the coherent superposition of two spectra with different frequencies and amplitudes will produce a superimposed spectrum with peaks and troughs. Figure 6 The simulation yielded two sets of final output spectra from the DFB laser chip, showing the peaks and valleys described above.

[0068] Simultaneously, once the reflectivity of the end-face coating is determined, the feedback intensity of the light-emitting end-face will also be determined, meaning the ripple intensity of its gain peak can be determined; the higher the reflectivity, the higher the ripple intensity. At this point, according to... Figure 6 The difference h between the highest and lowest points of the interference peak can be used to qualitatively determine the feedback intensity and amplitude intensity of the docking end face 2. The resonance spectrum and amplitude intensity of The cancellation of the peaks and troughs of the two peaks in the resonance spectrum decreases or their superposition increases. The greater the end-face feedback intensity, the greater the intensity difference h between the peaks and troughs, thus the parasitic feedback intensity of the docking end face can be determined.

[0069] The analysis of the spectrum tested by the spectrometer includes: 1) analyzing the spectral jitter caused by feedback from the light-emitting end face in the right side of the lasing wavelength (the wavelength region longer than the lasing wavelength). If a new spectral jitter is superimposed on the normal light-emitting end face feedback jitter Ripple, it is determined that the DFB laser chip has parasitic optical feedback at the docking end face; 2) by reading the intensity of the parasitic jitter of the Ripple caused by the parasitic optical feedback, i.e., the height difference h between the peak and trough of the superimposed waveform (reading the height difference directly in the spectrum), and comparing it with the different intensity difference h obtained from different batches of chips, the comparison relationship of the high and low intensity of the end face feedback parasitic intensity of different designs is finally obtained.

[0070] Example 2

[0071] A method for detecting the parasitic optical feedback intensity at the docking end face of a DFB laser chip in a docking structure, the specific operation steps of which are as follows:

[0072] (1) such as Figure 1As shown, based on the design of an Al-containing quantum well structure for a DFB laser chip, a primary epitaxial wafer was obtained, and two photomasks were drawn. Figure 1 One type of structure has no docking and the other has docking. The docking surfaces are etched using semiconductor processes such as photolithography, dry etching, and wet etching. To verify the method's ability to detect and identify parasitic feedback intensity, the docking surfaces are prepared at two different angles, as shown below. Figure 7 The angles of end face 1# are 63° and end face 2# are 75°. Finally, the final optical chip material structure is obtained by performing Al-free material docking and epitaxy. According to the law of reflection, the feedback intensity increases with the increase of the angle of the reflecting end face. Therefore, it can be judged that the parasitic feedback of end face 2# is stronger than that of end face 1#.

[0073] (2) Using semiconductor optical chip technology, ridge waveguide etching, electroplating, annealing, thinning and other processes are performed to obtain the process epitaxial wafer. The cleavage is carried out according to the cavity length design period (the wafer layout can be directly seen). In this embodiment, the laser cavity length non-docking comparison chip uses a cavity length of 150um and a chip width of 250um, while the docking test optical chip uses an active region of AlGaInAs material of 150um + passive InGaAsP docking regions of 15um at both ends. The center point of optical coating reflectivity is determined according to the grating lasing wavelength position. In this embodiment, the optical chips in the three different cases in Table 1 are obtained by processes such as grating lasing wavelength of 1286nm.

[0074] Table 1 Three types of optical chips

[0075]

[0076] (3) To facilitate the application of bias current and temperature control, three different chips were eutectic-coated onto a COC carrier and then mounted on a TEC for temperature control. Spectrometer testing was performed; in this embodiment, a Yokogawa 6370D spectrometer was used. The test type was High2, the test resolution was 0.02 nm, the test spectral width was 50 nm, the chip test temperature was 25°C, and the test current was 50 mA. The test spectra of the three chips are as follows: Figure 8 As shown, a is the spectral spectrum measured by spectral analysis, and b is the spectrum obtained after adjusting the range of values ​​on the coordinate axes. Figure 8 The conclusions in Table 2 can be drawn from this.

[0077] Table 2. Test and Analysis Conclusions of Three Optical Chips

[0078]

[0079] The test spectrum results and analysis conclusions show that the method of this invention can accurately determine whether there is parasitic optical feedback generated at the docking end face, and the level of this parasitic feedback can be effectively detected by reading the normalized height difference h in the test spectrum, thus providing effective support for the improvement of docking process and material design.

[0080] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the protection scope of the present invention.

Claims

1. A method for detecting the parasitic feedback intensity at the docking end face of a DFB laser chip in a docking structure, characterized in that, A spectrometer was used to perform high-precision spectral testing on the DFB laser chip to obtain sampling data. The jitter state caused by the different wavelength point light intensities of the obtained test spectral morphology was analyzed and calculated to determine whether there is internal end-face optical feedback in the docking structure. The relative intensity of the end-face feedback of the docking structure was analyzed to obtain the parasitic feedback intensity.

2. The method for detecting the parasitic feedback intensity at the docking end face of a DFB laser chip in a docking structure according to claim 1, characterized in that, Includes the following steps: Step S1: A DFB laser chip with docking structures on the optical end face and the back light end face is fabricated using active light-emitting region epitaxy and docking epitaxy; Step S2: The wafer of the prepared DFB laser chip is cleaved and sliced ​​to form a single laser chip, and the single laser chip is placed on the COC carrier by eutectic bonding and the positive and negative electrodes of the COC carrier are connected by gold wires. Step S3: Fix the COC carrier on the test stage, attach the positive and negative electrodes of the power probe to the corresponding positive and negative electrodes of the COC carrier, and simultaneously couple the light emitted by the single laser chip into the spectrometer using a wedge-shaped end-face coupling optical fiber at the light-emitting end face of the single laser chip. Step S4: Set the spectrometer test parameters, stabilize the COC carrier with DC power, and read the spectrum of a single laser chip; Step S5: Analyze and calculate the read spectrum to determine whether there is parasitic optical feedback at the docking end face; compare the Ripple intensity of the optical feedback spectrum of single laser chips from different design batches to obtain the comparison relationship of the intensity of parasitic optical feedback at the docking end face of different designs.

3. The method for detecting the parasitic feedback intensity at the docking end face of a DFB laser chip in a docking structure according to claim 2, characterized in that, The analysis of the test spectrum includes: 1) analyzing the spectral jitter caused by feedback from the light-emitting end face on the right side of the lasing wavelength in the spectrum. If a new spectral jitter is superimposed on the normal light-emitting end face feedback jitter Ripple ripple, it is determined that the DFB laser chip has end face parasitic optical feedback of the docking structure; 2) by reading the intensity of the parasitic jitter of the Ripple ripple caused by the parasitic optical feedback, i.e. the height difference h between the peak and trough of the superimposed waveform, and comparing it with the different intensity differences h obtained from different batches of chips, the comparison relationship of the high and low intensity of end face feedback parasitic intensity of different designs is finally obtained.

4. The method for detecting the parasitic feedback intensity at the docking end face of a DFB laser chip in a docking structure according to claim 3, characterized in that, Both the active light-emitting region epitaxy and the docking epitaxy are completed by metal-organic chemical vapor deposition; the active light-emitting region epitaxy adopts an AlGaInAs quantum well structure, and the docking structure adopts an InGaAsP quantum well structure.

5. The method for detecting the parasitic feedback intensity at the docking end face of a DFB laser chip in a docking structure according to claim 3 or 4, characterized in that, The bevel angle of the wedge-shaped end face coupled optical fiber is 8°±3°.

6. The method for detecting the parasitic feedback intensity at the docking end face of a DFB laser chip in a docking structure according to claim 5, characterized in that, The spectrometer test parameters include: spectral scanning sensitivity ≥ 20 seconds / 100 nm; spectral scanning accuracy ≥ 0.01 nm; and total spectral scanning width of lasing wavelength ± 20 nm to lasing wavelength ± 80 nm.

7. The method for detecting the parasitic feedback intensity at the docking end face of a DFB laser chip in a docking structure according to any one of claims 3, 4, and 6, characterized in that, A docking structure is provided at both ends of the active light-emitting region. A high-reflection coating is provided on the outer backlight surface of the docking structure on the backlight end face, and a low-reflection coating is provided on the outer side of the docking structure on the emission end face. The effective length of the optical resonant cavity formed by the active light-emitting region and the docking structure at the input end is L1, and the physical length of the DFB laser chip formed by the active light-emitting region and the two docking structures is L2. When light waves oscillate within a resonant cavity, they form standing waves. Only wavelengths that meet the resonance condition can exist stably within the optical waveguide and be output through the output end face of the DFB laser chip to form effective optical power. All light waves that can form stable resonance are selected by the grating to form the main lasing peak, while light waves that are not selected by the grating are emitted in the form of small gain peaks. The height of the small gain peaks formed by the feedback from the output end face is called the ripple.

8. The method for detecting the parasitic feedback intensity at the docking end face of a DFB laser chip in a docking structure according to claim 7, characterized in that, The longitudinal mode spacing d of the gain peak for: Where λ is the wavelength of the calculated interval frequency band, ñ is the effective refractive index of the optical waveguide, and L is the effective cavity length of the waveguide resonant cavity; The effective refractive index ñ is determined by the lasing wavelength λ of the grating lasing peak of the DFB laser chip. g =2ñ∧ is calculated, where ∧ is the grating period.

9. The method for detecting the parasitic feedback intensity at the docking end face of a DFB laser chip in a docking structure according to claim 8, characterized in that, The longitudinal mode spacing d of the spectral gain peak is obtained from the read spectrum. The effective cavity length L is calculated using the effective refractive index ñ. By comparing the effective cavity length L with the lengths of different chip structures, it can be determined which feedback end face is responsible for the gain peak. Furthermore, the grating period can be adjusted to make the grating lasing peak avoid the strongest and most obvious gain center, thus avoiding the impact on the readout accuracy of the Ripple ripple of the gain peak. The height of the gain peak is determined by the residual reflectivity of the low-reflectivity coating on the exit end face; The low-reflection coating on the light-emitting end face is of the same order of magnitude as the nearby docking structure, and the feedback from the docking structure near the light-emitting end face affects the spectral morphology.

10. The method for detecting the parasitic feedback intensity at the docking end face of a DFB laser chip in a docking structure according to claim 9, characterized in that, If there is no feedback at the cross-section of the docking structure, the output spectrum will only contain the grating lasing peak and a regular parabolic resonant spectrum. If there is parasitic small feedback at the end face of the docking structure, the optical resonant cavity, the active light-emitting region, and the resonant cavity composed of the two docking structures will each generate a resonant spectrum. Due to dual-spectral interference, in addition to the grating lasing peak, the output spectrum will also contain a resonant spectrum with high and low fluctuations, and the intensity of the grating lasing peak will be much greater than that of the two resonant spectra. According to the principle of two-peak interference, interference fringes will be formed between the two resonant spectra. The spectral amplitude intensity A after the superposition of the two interference spectra is: A1 A2 ];in, The amplitude intensity of the resonant spectrum of the resonant cavity including the backlight end-face contact structure. The amplitude intensity of the resonant spectrum of the resonant cavity, which includes the interface structure of the backlight end face and the light-emitting end face. The initial phase of a resonant spectrum, This is the initial phase of another resonant spectrum; After determining the reflectivity of the end-face coating of the docking structure, the height of the ripple of the gain peak is determined; the feedback intensity of the docking structure end face is determined based on the difference h between the highest and lowest points of the interference peak in the measured spectrum; amplitude intensity. The resonance spectrum and amplitude intensity The peaks and troughs of the two peaks in the resonance spectrum cancel each other out or overlap and increase. The greater the feedback intensity, the greater the difference h between the peaks and troughs, thus obtaining the parasitic feedback intensity of the docking end face.

Images