Patent Application
20220360038
The subject matter of this application relates to systems and methods for externally modulating light from a laser, such as an Electro-modulated laser (EML) where a laser and an external Electro-absorption Modulator (EAM) are fabricated as a single device, or other similar assembly.
Optical communications systems transmit information between devices by modulating light propagated through a fiber optic cable or some other optical medium. For example, many existing optical communications systems use a laser to produce a light beam having a narrow line width spectrum and provides a mechanism for modulating that light. The modulation of the light from the laser is what carries the information in the signal. The information-carrying light is then propagated to a receiver using a single mode fiber optic cable, which can transmit signals over longer distances relative to multimode fiber.
Light from a laser is typically modulated in one of two ways—direct modulation and external modulation. A directly modulated laser uses an information-carrying signal to drive the laser, thus the output from the laser carries the modulation signal. A directly modulated optical transmitter is a cost-effective solution for many applications, however it tends to produce a relatively large amount of laser chirp, which is an undesired change in the center frequency of the laser caused by the intensity modulation that produces the information-carrying signal. At the minimum loss wavelength of standard optical fiber, this laser chirp will interact with the optical dispersion inherent in the fiber optic cable, which is the tendency of spectral components of an optical signal to travel at different velocities along the fiber. Chirp also interacts with the Raman scattering that occurs in fiber to produce additional noise at the receiver. This is commonly referred to as Interferometric Intensity Noise (IIN).
The interaction of chirp with fiber dispersion produces undesirable performance degradations, such as composite second order (CSO) distortions. Though these distortions can be corrected using an electronic circuit that pre-distorts the input signal to the laser in a way that “cancels” the CSO distortion due to the chirp/dispersion in the fiber, so as to produce the original undistorted signal at the receiver, the distortion correction has to be customized for the transmission length of fiber run between the transmitter and receiver because fiber dispersion is a function of fiber length. Therefore, this requires additional tuning during network implementation. Also, this may cause some limitations in certain applications. For example, when the light is split in the transmission path and each portion of the split light travels down to different fiber lengths, it is difficult to design a predistortion correction circuit that suits both transmission lengths. Furthermore, when a primary link and a secondary link have different link lengths, the distortion correction needs to be reset after switching occurs between the primary and secondary links. Finally, electronic distortion correction has its own limit in terms of its correction capability, which limits the total transmission link length.
An externally modulated laser transmitter, conversely, has a laser produce an unmodulated output, which is then fed into an external circuit that modulates the output. There are different types of external modulator technologies, such as a lithium niobite (LN) based Mach-Zehnder (MZ) modulator and an electro-absorption modulator (EAM). For LN MZ transmitters, the light from the light source is split equally and each portion is sent to a phase modulator that uses a voltage to modulate the phase due to an electro-optic effect. The light from the two separate paths are then combined and interfere. If the phase difference between the two light beams is zero degrees, then the maximum optical output power is achieved. If the phase difference between the two light beams is 180 degrees, then the minimum optical output power is achieved. The LN MZ based external modulator thus provides very good analog performance over long transmission distance not only because its low modulator chirp, but also because so long as the LN MZ modulator is biased at its quadrature point, it produces very low second order distortions.
However, LN MZ transmitters also suffer some drawbacks. First, the best second order distortion performance can only be achieved at a quadrature point of the modulator transfer function, and a small bias deviation from that point makes the distortion degrade very quickly. Therefore, the modulator voltage bias for the optimal performance needs to be constantly monitored and controlled because to prevent drift. Also, the modulator is bulky and costly as compared to a directly modulated transmitter.
A typical electro-absorption modulator (EAM) relies on the Franz-Keldysh effect or Quantum-Confined Stark Effect (QCSE) where the effective band gap of the semiconductor changes in response to an applied voltage. At certain wavelengths, this change in bandgap causes a change in absorption. The change in absorption is used to selectively either absorb or pass the light from the laser. Any absorbed light is converted to photocurrent, and therefore the electro-absorption modulator (EAM) works in a similar way to that of a photodetector when the appropriate bias is applied. A particular EAM may allow one bias voltage to be applied that causes the modulator to absorb all, or substantially all of the incoming light at the laser's wavelength while another applied bias voltage allows the light at the laser's wavelength to pass through transparently. Thus, alternating the bias of the EAM may produce either analog or digital signals depending on the modulation scheme.
EAM external modulation of laser transmitters have several advantages. First, the electro-absorption modulator has a much lower chirp as compared to the directly modulated DFB laser. Second, the electro-absorption modulator requires a low bias voltage and driving power for modulation. Third, the electro-absorption modulator can be integrated with a DFB laser to form a device called an EML (electro-absorption modulated laser). Because of this integration, the EML device is very small with a package similar to a normal DFB laser, and therefore very cost effective.
The change in bandgap associated with the Franz-Keldysh or the QCSE is very fast. Therefore, EAMs that rely on these effects are commonly used for high speed digital communications. However, they have drawbacks for analog communications. The main drawback is they have a very non-linear absorption versus bias curve. This is acceptable and can even be preferred in digital applications but is not desired in analog applications where a linear change in absorption over a wide bias interval is preferable.
An alternate way to design and use an EAM is to operate at a wavelength that has significant absorption at zero volts bias and the minimum amount of light will pass. Applying a positive bias will cause the EAM to become more transparent as an approximately linear function of the positive bias voltage until full transparency is reached and the maximum amount of light will pass. However, the precise wavelength setting to achieve optimum performance as well as the method to set this wavelength is presently unknown. Conversely, the optimum design for the EAM active region to get optimum performance at a desired wavelength is unknown.
For optimum EAM performance, it is desired to minimize the light leakage at 0V bias. However, setting the wavelength too long or too short will potentially result in more light leakage at 0V bias, reducing EAM performance. Increasing the EAM length can reduce the amount of leakage but may not be desirable for a number of reasons.
What is desired, therefore, is an improved externally modulated laser that reduces light leakage without increasing the length of an EAM section.
For a better understanding of the invention, and to show how the same may be carried into effect, reference will now be made, by way of example, to the accompanying drawings, in which:
The external modulator 14 is operated at an electrical bias 18 which is denoted as BiasM. An information-containing electrical signal 20 that is modulated around a mean value of 0, denoted in
The modulator 14 and laser 12 are typically made from direct bandgap semiconductors such as Indium Phosphide (InP), Gallium Arsenide (GaAs) and/or related materials that exhibit direct bandgap properties. The laser is typically a Distributed Feedback (DFB) laser, which produces a narrow linewidth, single-mode optical output beneficial for long distance fiber-optic communications. However, it may consist of any laser implementation capable of producing a narrow linewidth output including, but not limited to a Distributed Bragg Reflection (DBR) laser or an External Cavity laser.
One preferred embodiment of the laser 12 and modulator 14 shown in
The active region 52 preferably comprises very thin, alternating layers 54a, 54b of low band-gap energy material that is in between material with higher band-gap energy. The way the conduction and valence bands align in this structure results in the creation of energy wells in both the conduction and valence bands. When the thickness of these energy wells is in the range is a few atoms thick to a few 10s of atoms thick, quantum mechanical effects dominate, and the energy wells are referred to as Quantum Wells (QWs). The energy states contained within the these QWs are no longer equivalent to the bulk material energy state and will instead depend on the energy depth of the wells and thickness of the QWs. There are generally material system limitations that limit the ability to control the depth of the QWs, but by careful control of the thicknesses, it is possible to control the bandgap energy of the active region. As explained below, because bandgap energy is related to the gain and absorption spectrum, it is possible to tailor the active region to produce a desired optical gain and/or absorption spectrum. The “band-gap” energy of a material refers to the amount of energy input required to cause an electron to rise from a valence state to a conductive state, and conversely the amount of energy an electron must emit or release when falling from a conductive state to a valence state. When the energy that causes the electron to rise to the conduction band comes from a photon, this is referred to as photo-absorption. When the energy that is emitted is in the form of a photon, this is referred to as photoemission.
When an electron, normally in a valence state, absorbs enough energy to rise to a conductive state, it can propagate through the material in response to an electric field. It also leaves behind a hole in the valence band that behaves similarly to positively charged particle that propagates in the opposite direction in response to the same electric field. The movement of electrical current through a semiconductor may therefore also be conceptualized as the as movement of negatively charged electrons going in one direction and positively charged holes going in the opposite direction.
Light with energy higher than the bandgap energy incident on a semiconductor will be absorbed by electrons in the valence band, causing them to jump to the higher energy conductive band state, thereby creating an electron and a hole pair. If an electric field is present, the electron and hole will separate and not recombine. In the case of an EAM at 0V bias, an electric field will be present due to the built-in potential of the PN junction. This provides the electric field that separates the electron and hole pair causing them to not recombine.
At some positive bias, the electron—hole pairs will not separate because the bias applied will overcome the built-in potential of the PN junction. This will allow the electron—hole pair to recombine and release their energy. In the case of direct bandgap semiconductors such as GaAs and InP, the energy is most often released in the form of a photon with an energy equal or approximately equal to the difference in energy states between the electron and hole. This emission of a photon can be stimulated by a passing photon causing a coherent addition to power. When the number of photons being absorbed approximately equals the numbers of photons being generated by stimulated emission, you have a condition known as transparency. At a bias higher than the bias needed to overcome the built-in potential of the PN junction, additional electrons and holes will be injected into the active region. This can result in a stimulated emission rate that exceeds the absorption rate. In this state, the EAM will produce gain and the power output from the EAM can exceed the power input.
The laser section also requires a bias sufficiently high to produce gain. However, the laser requires gain that is high enough to overcome losses. It also requires optical feedback from facets or other reflection sources to “seed” the stimulated emission and create the self-sustaining condition known as lasing. By designing a structure to provide feedback at a desired wavelength of light, such as a DFB structure, a narrow linewidth laser is generated. A laser may be created with the quantum well structure of the active region 52 of the laser section 56, so long as gain overcomes loss.
The structure shown in
For IL type EML devices, a trench 68 is commonly etched between the DFB laser section 56 and EAM section 58 down to the intrinsic SCH layer 60 to provide electrical isolation between the laser section 56 and EAM section 58. For BJC EML lasers, this isolation is achieved either by leaving a small gap or applying an electrically insulating coating to the front facet of the laser and/or the back facet of the EAM section. These coatings are also commonly designed to have anti-reflective properties, which can help reduce optical feedback from the interface between the EAM and laser section.
By selecting an appropriate bias voltage for the laser section and the EAM section, respectively, the QWs generate light in the laser section 56 and absorb light in the EAM section 58. As explained earlier, in the laser section 56, application of a sufficient forward (+) bias will generate stimulated emission of photons of a wavelength determined primarily by the properties of the reflector 62.
The EAM section has its bias modulated to either absorb or transmit light at the wavelength of that emitted by the laser. The modulation type adopted is typically either a negative bias modulation where the bias is modulated between 0V and some negative voltage, such as −1V, or a positive bias modulation where the bias is modulated between zero volts and a positive voltage such as +1V. These operations are illustrated in
Referring specifically to
Positive bias modulation, conversely, is typically used for analog applications such as optical RF-over-fiber. Referring to
Between 0V and the bias required to achieve transparency, the optical output power will be nearly linearly dependent on the current extracted from the EAM section. Furthermore, the current extracted from the EAM section will be largely linearly dependent upon the voltage applied to the EAM, achieving a nearly linear relationship between light output power and EAM input voltage. Because of this primarily linear relationship between EAM input voltage and optical output power, a modulated electrical signal that is applied to the EAM will be reproduced as an optical power modulation with low distortion, which is important for applications such as RF-over-fiber.
As noted, at 0V bias as shown in forward-biased implantation as seen in
The present inventor realized that the absorption/gain spectrum of a material in an active region exhibits a localized “exciton absorption peak,” which results from an artifact in the quantum well structure where electron-hole pairs are weakly bound together in the QW structure. This peak is observable at or near room temperature in the absorption spectrum of most QW and MQW active region designs. In a preferred embodiment, the operating wavelength 74 of the laser section 56 of an EML or laser and EAM system is preferably determined based on the exciton absorption peak of the material comprising the active region of the EAM, at room temperature or another ambient temperature in which the EAM is intended to operate (e.g., an ambient temperature surrounding a node or optical network unit in a communications network). In some preferred embodiments, the EAM may be operated in a forward bias manner. In another preferred embodiment, the operating wavelength 74 of the laser section 56 of an EML is set near or at the exciton absorption peak at zero volts. Notably, as can be seen in
This localized peak represents a maximum absorption coefficient of the active region material for any given length of an EAM, as can be seen in
There may be more than 1 exciton peak present in the absorption spectrum at or near room temperature (or other ambient temperature for the EML) depending on the QW or MQW active region structure. Multiple peaks are caused by the presence of carriers with different effective masses, the most common of which are light holes and heavy holes. The excitons formed by light holes and heavy holes may have different binding energies, which results in slightly different absorption peak wavelengths. In the event of multiple exciton peaks in the absorption spectrum, the lasing wavelength can be set to the peak that is at the shortest wavelength, the exciton peak closest to the absorption band edge or any peak in-between, provided the condition of transparency or near transparency can be obtained at that wavelength under positive bias.
At the forward bias required to achieve transparency, the extracted current will be at or near 0 and the optical output power will be equal or nearly equal to the optical input power. Further increase of the forward bias will result in current injection into the active region of the EAM, which may result in optical gain. This gain can result in the optical output power from the EAM section exceeding the optical input power. The increase in optical output power will be primarily related to the injected current until the gain saturates. In this manner, linear or near linear operation can be extended to a maximum optical output power that exceeds the optical input power to the EAM section.
Although the foregoing description used an EML as an exemplary device by which to illustrate the systems and methods disclosed in the present application, those of ordinary skill in the art will appreciate that these systems and methods may be used in other arrangements. As one example, the disclosed systems and methods may be used in an arrangement where a laser and an EAM are separately fabricated, and the laser supplies light to the EAM via an optical fiber or other transmission medium such as air, and the EAM modulates that light.
It will also be appreciated that the invention is not restricted to the particular embodiment that has been described, and that variations may be made therein without departing from the scope of the invention as defined in the appended claims, as interpreted in accordance with principles of prevailing law, including the doctrine of equivalents or any other principle that enlarges the enforceable scope of a claim beyond its literal scope. Unless the context indicates otherwise, a reference in a claim to the number of instances of an element, be it a reference to one instance or more than one instance, requires at least the stated number of instances of the element but is not intended to exclude from the scope of the claim a structure or method having more instances of that element than stated. The word “comprise” or a derivative thereof, when used in a claim, is used in a nonexclusive sense that is not intended to exclude the presence of other elements or steps in a claimed structure or method.
This application claims the benefit of U.S. Provisional Patent Application Ser. No. 63/184,452 filed May 5, 2022.
| Number | Date | Country | |
|---|---|---|---|
| 63184452 | May 2021 | US |
