Recently, investigations of gravity waves have seen growing interest, because this phenomenon is important for the coupling of several atmospheric layers and, therefore, for the overall effect on atmospheric dynamics [1–5]. Observations of gravity waves are especially interesting in the mesosphere and lower thermosphere (MLT), where breaking and, thus, energy and momentum transfer occurs [1,2]. Temperature and wind distributions in this atmospheric region can be used to obtain data of temporal and spatial characteristics of gravity waves. Measuring the Doppler broadening and shift of a resonance line of metal atoms in the MLT by means of resonance lidar systems is one well-established and reliable approach for obtaining such temperature data with high temporal and spatial resolution [2,4,6]. As presented in a former publication of our group [7], where additional details about applications and state of the art in Alexandrite laser technology, in general, are presented, diode-pumped $Q$-switched Alexandrite lasers in single longitudinal mode (SLM) operation are a suitable laser source for potassium resonance lidar systems [8,9]. Directly addressing the potassium resonance (770 nm) with the fundamental wavelength or the iron resonance (386 nm or 372 nm) with only one conversion step allows the design of robust and compact lasers.

A laser source deployed in a resonance lidar is especially challenging because it requires a narrow linewidth with SLM operation and high spectral purity [9]. When a laser with a linear resonator operates in SLM, spatial hole burning significantly reduces the laser efficiency which can completely prevent SLM operation. The twisted mode technique [10] which prevents spatial hole burning in a linear resonator cannot be used due to the birefringence of Alexandrite. Therefore, a travelling wave ring resonator is developed for SLM operation.

First atmospheric measurements of a resonance lidar system with a diode-pumped Alexandrite ring laser were successfully conducted and published in Ref. [7]. The laser used for these measurements (DALI-1) could already achieve all parameters required for measuring the Doppler shift of the resonance line of metal ions in the MLT as well as the Doppler shift of aerosols in lower atmospheric regions. A summary of the required parameters is given in Table 1. The details regarding the conducted measurements, as well as the respective lidar technology, will be published soon in an additional paper. The more advanced DALI-2 laser presented in this publication is an improved version of the DALI-1 laser, using a new resonator design enabled by a novel pump scheme. The progress in laser design and the successful atmospheric measurements highlight the potential of diode-pumped Alexandrite laser technology for advanced lidar sources.

Table 1. Requirements for the Beam Source in a Potassium Lidar System and Achieved Parameters for the Alexandrite Laser

View Table

The DALI-1 laser is pumped by two commercially available laser diode modules emitting at 636 nm. Detailed information on the modules can be found in Ref. [7]. The beam quality of the diode modules is ${\mathrm{M}}^2=30$ in fast-axis and ${\mathrm{M}}^2=300$ in slow-axis for each diode module, and is unaffected by a change in a repetition rate or pump pulse duration. The considerable asymmetric beam quality in fast- and slow-axis results in an elliptical effective pump beam size. Consequently, different focal lengths in the two directions of the resulting thermal lens are induced, making it challenging to design a stable and robust resonator. In order to overcome these drawbacks of the former design, the two pump modules used for DALI-2, which are of the same type as the ones used in DALI-1, are polarization-coupled and the fast- and slow-axis beam qualities are symmetrized by means of a step mirror [11] to provide a round pump beam cross section.

Figure 1 shows a schematic view of the pumping scheme. The pump light of the two modules is linearly polarized and the polarization is oriented perpendicular to each other by means of half-wave plates. The beam of each module is magnified by a factor of three in slow-axis with a cylindrical telescope and polarization-coupled with a polarizing beam combiner cube. Afterwards, the combined beams are rearranged by means of a step mirror and the resulting beam is magnified by a factor of three in fast-axis before being focused into the 7 mm long Alexandrite crystal. The combined pulse energy behind all pump optics is 24 mJ at a pump pulse duration of 120 μs, with a total transmission efficiency of the complete beam shaping optics of $T=0.85$.

 

Fig. 1. Schematic drawing of the pump beam shaping.

Download Full Size | PDF

From the caustic measured behind all pump optics, one can derive a beam quality of ${\mathrm{M}}^2=100$ in fast-axis and ${\mathrm{M}}^2=150$ in slow-axis. The focus radii are $w=162\mathrm{\mu m}$ in fast- and $w=237\mathrm{\mu m}$ in slow-axis. Taking into account the caustic of the pump light weighted with the absorption in the crystal, one can calculate an effective pump beam radius in the crystal, which is 215 μm in fast- and 276 μm in slow-axis. A section of the measured caustic is plotted in Fig. 2. Due to the refractive index of Alexandrite being 1.77 for the pump wavelength, the divergence of the pump beam is changed and, therefore, the optical length for the pump light is shortened which results in an optical length of the crystal of approximately 4 mm.

 

Fig. 2. Development of the beam radius of the pump light along the optical axis with a beam profile at the focus position (inlet). The calculated effective beam radius inside the crystal is 215 μm in fast- and 276 μm in slow-axis.

Download Full Size | PDF

After polarization-coupling of the pump modules, the resulting polarization of the pump light is 50% horizontal and 50% vertical in reference to the b-axis of the Alexandrite crystal, which is cut along the c-axis and has a dopant concentration of 0.29 at%. The Alexandrite crystal is operated at a temperature of 105°C. At its single pass through the laser crystal, 40% of the pump light is not absorbed, since the absorption at 636 nm is a magnitude lower for light vertically polarized to the b-axis of the crystal [12]. The residual pump beam is collimated behind the laser crystal, reduced in size by a 1:4 telescope, reflected by a plane mirror and refocused into the laser crystal. To increase the absorption of the light on the return trip through the crystal, the polarization is rotated by double-passing through a quarter-wave plate.

The laser resonator is schematically shown in Fig. 3. It has a total length of approximately 2000 mm and consists of two dichroic pumping mirrors with high transmission for the pump wavelength, two concave mirrors with a radius of curvature of $-900\mathrm{mm}$, and several plane folding mirrors including the output coupler having a reflectance of 0.97 for the laser wavelength. To actively stabilize the cavity length, one folding mirror is mounted on a piezo stack. For $Q$-switching operation, a Brewster-angled Pockels cell and two thin-film polarizers are included in the cavity with the polarization being adapted accordingly by two half-wave plates. The unidirectional operation is ensured by a Faraday rotator.

 

Fig. 3. Schematic drawing of the resonator.

Download Full Size | PDF

The resonator is designed for a laser beam radius of around 200 μm in the laser crystal, which results in a good overlap with the pump beam. The beam radius on critical optical components such as the Pockels cell or the Faraday rotator is around 600 μm, preventing laser-induced damage.

In $Q$-switched operation the laser yields a pulse energy of 1.7 mJ and features a high pulse-to-pulse stability with a standard deviation of 0.2% at a repetition rate of 500 Hz, resulting in an average power of 0.85 W. A measurement of the pulse energy over time is plotted in Fig. 4. The high temporal stability of the pulse energy is a consequence of the much more robust resonator design, but is not necessary for the lidar measurements, since the pulse energy of each pulse is monitored and considered in the calculation of the received signal. The low gain of the laser medium results in a pulse duration of the $Q$-switched pulses of 850 ns. The polarization of the output beam is adjusted with a quarter-wave and a half-wave retardation plate without energy loss. The laser features an electro-optical efficiency of 2% when both the electro-optical efficiency of the diode modules of 33% and the transmission losses of the pump optics of approximately 15% are taken into account.

 

Fig. 4. Measured pulse energy in $Q$-switched operation over more than 30 min with a zoom on the relevant energy range as the inlet.

Download Full Size | PDF

The optical design of the resonator also allows an operation with a repetition rate of 250 Hz. In this case, the radius of curvature of the two concave mirrors is changed to $-1000\mathrm{mm}$. The pulse energy, spatial parameters, and SLM operation are comparable to the operation with 500 Hz.

The output beam is round and stigmatic so that no additional cylindrical beam shaping is necessary. The beam profile of the laser has a Gaussian shape with a beam quality of ${\mathrm{M}}^2\le 1.1$ in both spatial directions. A caustic of the output beam with beam profiles at selected positions of the optical axis is plotted in Fig. 5.

 

Fig. 5. Beam radius of the output beam plotted against the propagation length with the inset of the intensity profile at designated positions. The resulting beam qualities are ${\mathrm{M}}^2=1.10$ in the x-direction and ${\mathrm{M}}^2=1.06$ in the y-direction.

Download Full Size | PDF

SLM operation at 770 nm is achieved when the resonator is seeded through the output coupler with a commercial DFB diode laser in SLM operation with a linewidth of 1 MHz while the cavity length is actively stabilized. The linearly polarized seed laser beam is matched in its divergence and beam radius to those of the resonator mode. The average power of the seed laser is up to 17 mW, however, SLM could be achieved, even with one hundredth of this power without any change of the output parameters of the Alexandrite laser. A detailed description of the ramp-and-fire method used to stabilize the cavity length and the respective electronics can be found in Ref. [13].

Seeding the laser and stabilizing the cavity length result in a slight loss of pulse energy of 5%, while the spatial and temporal parameters of the Alexandrite laser are unchanged compared to the unseeded laser. The linewidth is measured with a spectrum analyzer developed at the IAP in Kühlungsborn. A comparison with the measured linewidth of the seed laser and the laser published in Ref. [7] indicates a linewidth $<10\mathrm{MHz}$. The frequency shift and jitter will be measured during a future measurement of the potassium layer in the atmosphere. The measured linewidth is stable over several hours without any alignment. A summary of the achieved parameters is shown in Table 1.

In conclusion, we demonstrated a $Q$-switched Alexandrite laser in SLM operation with Watt-level output power with an advanced pump concept based on symmetrized diode modules. The laser yields a pulse energy of 1.7 mJ at a repetition rate of 500 Hz and features an electrical-optical efficiency of 2%. When the resonator is seeded with a narrowband diode laser, SLM operation with a linewidth below 10 MHz is achieved.

The laser’s average power is more than five times higher, and the electro-optical efficiency is doubled compared to the previously published diode-pumped Alexandrite laser DALI-1, which was already used for atmospheric measurements [7]. Additionally, the laser is less sensitive to changes of the thermal lens or misalignment compared to its predecessor, thanks to the symmetric pump beam and careful laser design, as well as the usage of a single crystal instead of two. The rise in average power yields a higher resonance signal that shortens the integration time during a lidar measurement and, therefore, increases the temporal resolution. In comparison with the flashlamp-pumped SLM Alexandrite lasers in Refs. [8,9], the demonstrated average power and linewidth of the DALI-2 laser is in the same magnitude, while the electro-optical efficiency is two magnitudes higher. Atmospheric measurements with the DALI-2 laser will soon be performed to validate the applicability of the laser as a beam source in a resonance lidar system.