Construction of a Wavelength-tunable External Cavity Laser

Close up picture of the superluminescent diode use as a light source for this experiment

Abstract

We constructed a wavelength-tunable external cavity diode laser from a superluminescent diode and a diffraction grating. A diffraction grating placed in the feedback branch of the laser allowed for tuning the wavelength over a range of 824nm - 836nm. A spectrometer was used to measure the output spectrum, and from this we were able to measure the laser threshold along with the 1/3 nm mode spacing of the laser. A linear polarizer was used to measure the polarization of the laser, we express the elliptical polarization of the laser using a Jones vector. The Laser with feedback present has a Jones vector

Introduction

Laser short for "light amplification by stimulated emission of radiation" is a light source that has high temporal and spatial coherence. The spatial coherence means that the source is a very concentrated light source, and the temporal coherence means that the light is highly monochromatic. There are three main components to any laser. A gain medium, a power source, and an optical feedback system. The gain medium is the medium that emits the photons, and is therefore the source of light. the power source supplies the power which excites electrons in the gain medium, our laser uses a p-n junction and a current controller to supply current across the gain medium which excites electrons. Lastly, a laser needs an optical feedback system. This is really just some way of getting photons that left the gain medium to re-enter the gain medium to possibly interact with excited electrons. The most common way of ding this, and the way that we will do this, is with the use of two mirrors on either end of the gain medium. The two mirror will reflect photons back and forth thousands of times per second, which give the photons ample opportunity to induce stimulated emission in the gain medium. A diagram of the setup with feedback present is shown below.

Theory

Emission

In 1917 Albert Einstein released a paper called "Zur Quantentheorie der Strahlung", translated to "On the quantum theory of radiation", and in this paper he proposed the an excited electron could decay to it's ground state if an incident photon were to interact with the electron. This process is called stimulated emission, and is the basis for the theory behind lasers. There are two ways that a photon can be produced when an electron decays from an excited state to a ground state. the first is by spontaneous emission, as the names implies this emission is spontaneous, and the rate of spontaneous emission is dictated by the Einstein coefficients of the excited state. The other process is the very one that was outlined by Einstein in his paper, stimulated emission has one key advantage over spontaneous emission, and it is that the photon that is produced by stimulated emission is coherent with the incident photon. The process of spontaneous and stimulate emission are shown below.

Feedback and Boundary conditions

By providing feedback into the system, photons will travel between the external mirror and the back mirror in the SLD many times per second. The photons that re-enter the cavity have a chance to interact with electrons in the gain medium, and enhance stimulated emission in the SLD cavity. At some ``critical" injection current, for us around 100 mA, there are enough electrons in the conduction band, and enough photons travelling through the gain medium to maintain a steady supply of electrons for ``sustained stimulated emission". The transition to sustained stimulated emission is the defining characteristic of a laser.

Once we have sustained stimulated emission almost all of the photons in the gain medium will have the same phase and wavelength, we say ``almost all" because there will still be a small amount of spontaneous emission. Since the electric field must vanish at the boundaries of the cavity, we will see standing waves form in the SLD cavity. These standing waves must satisfy

Where n is a integer number, lambda is the wavelength of the photons, and mu is the index of refraction of the medium.

Laser Modes and doppler Broadening

The laser is operated at a fixed temperature, and therefore the gain medium will have thermal energy. Similar to how the velocity of air molecules in a room exists on a distribution, so will the energies of each electron in the conductance band. The distribution of energies has the effect of broadening the allowed energy gaps between the valence band and the conductance band. We see this spread in the emission spectrum and it is called ``Doppler broadening". The Doppler broadening of the output spectrum is dependent on the properties of the materials, which cannot be changed, and the temperature of the material, which can be. We use a temperature controller to maintain constant temperature.

The combination of boundary conditions, which yield cavity modes, and the gain curve results in laser modes.

The gain curve acts as a multiplier to the cavity modes. The cavity modes that exist underneath the gain curve are able to result in laser modes. There is a slight nuance here called the "laser threshold". The laser threshold is the threshold above which laser modes can occur, and below whiich there will not be laser modes. This can be though of using the analogy of fuel for a fire. There needs to be enough excited electrons at a certain wavelength to be able to create the positive feedback loop that results in a laser mode in that wavelength.

Everything up until now has to do with the basics of a laser, and how we are able to achieve "Lasing". Now we move on to talk about the process of making the laser "wavelength-tunable". We will also discuss the mathematical description used for analyzing the polarization, Jones Vectors.

Diffraction grating

The Blazed diffraction grating is used to separate a spectrum into discrete diffraction modes. The blazed diffraction grating is shown below.

The diffraction modes that result follow the formula.

Using the m=1 diffraction order, we get that the angle of diffraction. The m=2 diffraction order can be used for reflection, but this would not be as efficient as most of the light is contained in the m=1 diffraction order.

Polarization

Jones vectors are a way of describing the polarization of light in an optical system. The polarization is a vector, and the optical systems are matrices which act on the vector. We describe the vectors as a superposition of 2 orthogonal vectors. These vectors can be in phase or out of phase, and the amplitudes are described by the Jones vector.

If the phase difference is zero, we get Linearly polarized light, and if the phases are 90 degrees out of phase, and a_x=a_y, then we get circularly polarized light. To obtain elliptically polarized light, we need only to make the relative amplitudes not unity.

Apparatus

Basic Setup

The light source used in this experiment was a Hamamatsu L12856-04 SLD, and and was housed inside of the laser diode mount shown in the top right of the picture below. A constant current controller was used to provide the SLD up to 130 mA of injection current, which was a limit set to avoid damage to the SLD. Keeping the temperature of the SLD stable is crucial for this experiments, so a thermoelectric cooler was used to avoid shifts in the output spectrum.

A major difference between monolithic laser diodes and SLDs is that an SLD does not intrinsically create optical feedback for itself. Laser diodes, on the other hand, are fabricated in such a way that a highly reflective optical cavity is formed within the PN junction of the diode[1]. To make our SLD operate as a laser, we assembled an external cavity using a gold mirror in conjunction with the reflective back facet of the SLD. The external mirror is shown in the bottom left of the above picture. Lastly, our setup used a beamsplitter to reflect a portion of the light out of the cavity to where it can be measured by either a spectrometer or an optical power meter. A microscope slide is used the the above picture as a beamsplitter, while a non-polarizing beamsplitting cube is being used in the figure below.

Tuning Setup

The above figure shows an extension of the previous apparatus in which the laser was made tunable by use of a diffraction grating in the Littman-Metcalf Configuration. The diffraction grating is in the bottom left of the above picture. When light from the SLD hits the grating the first order diffracted light is reflected into a mirror on a rotation stage, where single wavelength is the reflected back into the grating and then the SLD. Light wavelengths higher and lower than what the setup is currently tuned to will be ejected from the system and not cause stimulated emission within the SLD.

In a final part of the experimental, a linear polarizer was placed in front of the optical power meter shown in the basic non-tuning setup. By rotating the linear polarizer, the two orthogonal polarization components of light produced by the setup can be measured.

Results

Output Spectrum and Power

With the addition of optical feedback provided by the externally mounted mirror, the occurrence of stimulated emission within the SLD was amplified. This resulted in increased output power at injection currents greater than 100 mA when compared to a setup without optical feedback. The comparison between the two setups are shown below, where their power curves diverge at approximately 100 mA.

When viewing the output spectrum at injection currents greater than 100 mA, sharp wavelength peaks begin to appear. These peaks are normally spaced about 1/3 nm apart, and exist in the 830-840 nm range. The specific location of these peaks varied with the components used in the experimental apparatus as well as injection current. The figure below shows the output spectrum using a non-polarizing beam splitter (NPBS) as an output branch from the setup, producing the usual 1/3 nm mode spacing and exhibiting several distinct wavelength peaks. An alternative setup using a microscope slide instead of the NPBS showed mode spacing of about 1 nm.

Wavelength Tuning

The laser was made tunable by use of a diffraction grating in the configuration shown in the Apparatus section. The animation below shows the output wavelength being tuned over the 824-836 nm region by rotation of a mirror in the feedback loop. The sensitivity was found to be 6.9 +/- 1.9 nm/degree, which is twice the angular dispersion of the diffraction grating and is not dependent on cavity length.

Polarization

The polarization characteristics of the laser were investigated by passing the output beam through a linear polarizer. By comparing the power output measured using different orientations of the linear polarizer, the primary and orthogonal field components can be revealed. The figure below shows that the SLD produces light that is nearly linearly polarized. The polarization extinction ratio, or ratio between the principal (90/270 degrees) and orthogonal (0/180 degrees) components of the electromagnetic field was calculated to be 13.8 dB when provided with optical feedback. The polarization can also be described by the normalized Jones vector

, which shows the relative magnitudes of the principal and orthogonal components. The orthogonal component is purely imaginary, since it is 90 degrees out of phase from the principal component. The +/- is show because it was not determined whether the the polarization was left or right handed.

Visual inspection of the SLD revealed that the output light was polarized (nearly) linearly in the direction orthogonal to the current path. In the figure below, current travels from the post labelled "SLD input", through a bond wire to the diode labeled "SLD", and finally to ground. The light output from the diode was hypothesized to be linearly polarized along the y axis, and was verified experimentally use the linear polarizer method previously stated.

Conclusion

A wavelength-tunable external cavity laser was constructed using a superluminescent diode as a light source. Optical feedback was provided using an externally mounted mirror and caused the occurrence of stimulated emission to become enhanced at a diode injection current of approximately 100 mA. Wavelength peaks were observed with a mode spacing of approximately 1/3 nm. The laser was made tunable using a diffraction grating in the Littman-Metcalf configuration, which allowed tuning over wavelengths from 824-836 nm. Lastly, the polarization of the laser was found to be highly elliptical with a polarization extinction ration of 13.8 dB.

Citations

[1] What is superluminescent diode? (SLD). FiberLabs Inc. "https://www.fiberlabs.com/glossary/about-sld/". (Retrieved May, 2021)