1 Introduction

Tunable lasers provide opportunity to adjust laser emission wavelength in a considerable range freely, and this flexibility of wavelength selection is desirable by many applications including spectroscopy, biomedical imaging, optical communications, frequency metrology, and interferometry [1, 2]. Among the known solid-state gain media, crystals such as Ti:Sapphire [3], Cr:Colquiriites (Cr:LiSAF, Cr:LiCAF, Cr:LiSGaF) [4], Alexandrite [5], Yb:Lu2O3 [6], Cr:Forsterite [7], Cr:YAG [8], Co:MgF2 [9], Tm:YLF [10], Cr:ZnS/ZnSe [11, 12] and Fe:ZnSe [13] offer particularly broad wavelength tuning ranges [14]. Via nonlinear interactions such as second- and third-harmonic generation and optical parametric amplification the working range of these systems could be further extended into visible and far-infrared regions [15].

Among the broadly tunable gain media, very few provide the advantages that Cr:LiSAF possesses [16,17,18]. Cr:LiSAF gain medium offers one of the widest emission bands in the near infrared region that manifests itself with an ultrabroad tuning range covering the 770–1110 nm interval [19,20,21] (a broader fractional tuning range is provided only by Ti:Sapphire: 660–1180 nm [22], and Cr:ZnS/ZnSe in the mid-infrared). Upper state fluorescence lifetime (τf) of Cr:LiSAF (67 µs) is around 20 times higher compared to Ti:Sapphire (3.2 µs) enabling flexibility in amplifier design, especially in terms of the requirement for pump sources. On the other hand, peak emission cross-section (σe) value of Cr:LiSAF is around 8 times lower than Ti:Sapphire. Overall, the product of τf and σe for Cr:LiSAF is around 2.5 times higher than that of Ti:Sapphire [23], and this provides a higher small signal gain to Cr:LiSAF. As an additional benefit, Cr:LiSAF crystals could be grown with very high optical quality with passive losses below 0.15% per cm [24, 25]. This enables construction of high-Q-laser cavities with record low (mW level) lasing thresholds [26]. Moreover, the full-tuning range of Cr:LiSAF could be scanned even in systems pumped with only 100 mW of pump power [25]. For comparison, reaching the 1000 nm wavelength region requires pump powers of around 5 W in Ti:Sapphire lasers systems due to lower gain and higher passive losses of the crystal [27, 28].

Despite the above mentioned advantages of Cr:LiSAF, Ti:Sapphire still dominates the laser research, and this is mostly due to weak thermomechanical strength of Cr:LiSAF. Thermal conductivity of LiSAF host is around an order of magnitude smaller than that of sapphire [4]. Despite its lower quantum defect, fractional thermal load could be quite high in Cr:LiSAF because of the presence of other undesired process such as Auger upconversion, excited state absorption, and thermal quenching of fluorescence lifetime [4]. Combined with its low fracture toughness [4], power scaling becomes rather challenging in Cr:LiSAF [16], even in thin disk geometry [29]. As a result, Cr:LiSAF has pros and cons as any other laser material, and for applications that do not require high average power levels, Cr:LiSAF-based systems could provide one of the best solutions.

As we have briefly mentioned, one of the most important advantages of Cr:LiSAF is its superior crystal quality: crystals with 10–20 times lower passive losses than Ti:Sapphire could be routinely grown [24, 25, 30]. As a result, even a Cr:LiSAF laser system pumped with a simple 100 mW pump diode could store up to 40 W of intracavity power (40 mW output with 0.1% output coupling) [25]. For comparison, at a pump power of 5 W, a typical Ti:Sapphire system produces around 1 W of output power at 1% output coupling, and hence only stores around 100 W of intracavity power. Thus, intracavity power storage ability of Cr:LiSAF is around 20-fold better compared to Ti:Sapphire. The large intracavity powers (intracavity photon numbers) are beneficial for minimizing laser noise, and simple Cr:LiSAF laser systems with sub-100-attosecond level timing jitter noise has already been demonstrated [31]. The large intracavity powers could also be used for efficient intracavity nonlinear experiments such as second-harmonic generation and optical parametric oscillation [32]. Since, extracavity nonlinear experiments require more complex systems with already higher peak power/energy levels, these systems are based on lasers with pulsed output as the peak powers of the cw systems are usually not high enough to drive nonlinearities in an efficient way. Hence, intracavity experiments could provide an advantage in terms of cost reduction, and could make these tunable laser sources more reachable to the laser community.

Ideally, by way of second-harmonic generation, the tuning range of Cr:LiSAF (770–1110 nm) may be extended into violet, blue, cyan and green regions of the spectrum: 385–555 nm. This wavelength range is interesting for many applications including spectroscopy, biomedical imaging, atom cooling/trapping, and quantum optics. The second-harmonic generation studies of Cr:LiSAF could be divided into extracavity and intracavity based efforts. In one of the earlier extracavity experiments, Pinto et al. used a 100 mJ Q-switched Cr:LiSAF laser system with 20-ns pulses to generate tunable second-harmonic (390–480 nm) and third-harmonic (260–320 nm) light using lithium triborate (LiB3O5, LBO) and BBO crystals [33, 34]. Second-harmonic and third-harmonic energies up to 20 mJ and 6 mJ were obtained in a 1 Hz system (average power levels: 20 mW) [34]. As another extracavity scheme, high peak powers from mode-locked Cr:LiSAF lasers were used to generate femtosecond blue pulses [35,36,37]. Agate et al. employed the output of a mode-locked Cr:LiSAF laser and used a 3 mm long potassium niobate (KNbO3) crystal, to generate 550 fs long blue pulses around 429 nm with 12.5 mW average power at 330 MHz [35].

Most of the earlier second-harmonic generation studies with Cr:LiSAF focused on intracavity scheme [16, 38,39,40,41,42,43], due to the above mentioned advantages. In 1995, using a 5 mm thick KNbO3 crystal, Falcoz et al. achieved cw blue powers up to 14 mW in the 427–443 nm range using a Cr:LiSAF laser pumped by two 500 mW multimode diodes (optical-to-optical conversion efficiency: 1.4%) [38]. Later in 1997, Laperle et al. produced up to 20 mW of cw blue power and a tuning range of 432–442 nm with LBO using a Cr:LiSAF laser pumped with up to 4 W of pump power (a fiber bundle pump system with 19 single emitter diodes) [39]. Via operating the system in quasi-cw mode, the second-harmonic power could be scaled up to 250 mW (25% duty cycle operation, 6.25% optical-to-optical conversion) [39]. In microlaser geometry, Eichenholz et al. realized 0.35 mW of output power at 430 nm from a 350 mW diode-pumped system using potassium niobate as the nonlinear crystal [40]. Using 1 W level multimode diodes, Makio et al. achieved single-line blue operation around 430 nm with up to 32 mW output power using LBO crystal (up to 120 mW of blue power is achieved in multi-line regime) [41,42,43]. Using a rather complex Cr:LiSAF laser pumped by four 1.8 W multimode diodes, Demirbas et al. obtained up to 1.16 W of second-harmonic power and a tuning range between 387 and 463 nm using BBO crystals (16% optical-to-optical conversion efficiency) [16].

All these previous nonlinear intracavity frequency generation experiments with Cr:LiSAF have used relatively complex multimode pump sources, and at the pump powers that are used, cooling of the diode/s as well as the laser crystal is required. Hence, the multimode diode pumping approach prevents the development of compact, portable and efficient laser system [35,36,37]. The focus of earlier work on usage of multimode diodes had been due to the limited output powers of red wavelength single-mode diodes. Initial single-mode diodes in red wavelength region provided only around 10 mW in early 1990s [44], and around 50 mW in early 2000s. The power levels scaled above 100 mW level in 2010 [25, 45] and to above 200 mW level few years ago. Hence, the recent progress in diode-technology opened up a way to reduce the complexity of tunable visible sources.

Motivated by this, in this current work, we have focused on using state-of-the-art 660 nm single-mode diodes with 210 mW output power as the pump source. These diodes do not require active cooling, can be driven by simple batteries and enable construction of low-cost laser systems with a very compact footprint. In the experiments, two such SMDs were used to first characterize the lasing properties of Cr:LiSAF in the infrared. Using 420 mW of pump power, we have achieved cw powers up to 190 mW, reached an efficiency up to 54% and acquired a tuning range from 795 to 1103 nm. In the frequency-doubling experiments, we have obtained up to 17.5 mW of output power at 425 nm, with a corresponding optical-to-optical conversion efficiency of 4.4%. Using six different BBO crystals with cut angles optimized in the 800–1000 nm range, a record frequency-doubling tuning interval of 402–525 nm is also achieved. To our knowledge, these are the first lasing results reported with Cr:LiSAF using the recently available 210 mW single-mode diodes. Moreover, the achieved cw frequency-doubled tuning range (402–525 nm) extends the earlier literature with Cr:LiSAF (387–463 nm [16]) considerably into the cyan and green regions of the spectrum. The obtained optical-to-optical conversion efficiency of the system (4.4%) proves the effectiveness of recent SMD diode pump sources. We believe that, this low-cost and compact laser system with ultrabroad tuning behavior in the near-infrared and visible regions of the spectrum could become useful for many scientific and technological application areas.

The paper is organized as follows: Sect. 2 describes the experimental setup. In Sect. 3, we present the cw lasing and tuning performance of the system in the near infrared and visible. Lastly, in Sect. 4, we summarize the results and provide a short outlook.

2 Experimental setup

Figure 1a shows a schematic of the diode-pumped Cr:LiSAF laser cavity used in cw laser experiments. Two linearly polarized, 660 ± 2 nm AlGaInP single-spatial-mode diodes (SMDs) with 210 mW output power are employed as the pump source. The electrical-to-optical conversion efficiency of the diodes are ∼33% and active cooling of the didoes is not required (diode operation range is specified as − 10 to 75 °C). The outputs of the diodes are first collimated by aspheric lenses (f = 4.5 mm) and then two 75-mm focal length lenses are used to focus the pump beams inside the Cr:LiSAF crystal. The 1.5% Cr-doped, Brewster-cut Cr:LiSAF gain medium is 10 mm long, and has an apertures size of 2 mm (thickness) by 10 mm (width). The electric field of the pump/laser light is TM (p) polarized, and is parallel to the c-axis of the crystal to maximize absorption and gain. The crystal absorbs more than 98% of the 660 nm pump light for the TM polarization. At these absorbed pump power levels, the Cr:LiSAF crystal also does not require active cooling, enabling the construction of quite simple and compact laser system.

Fig. 1
figure 1

(Color online) Schematic of the a cw b cw intracavity frequency-doubled Cr:LiSAF laser. The x-cavity is end-pumped by two 210 mW single-spatial-mode diodes (SMDs) at 660 nm. BRF: Birefringent filter

Full size image

An astigmatically compensated, x-folded laser cavity with two curved pump mirrors (M1 and M2, R = 75 mm), a flat end mirror (M3), and a flat output coupler (OC) is constructed. A long cavity arm length of ∼29 cm is used to obtain a beam waist of ∼25 ×∼40 μm inside the crystal. Two different custom-designed pump mirrors are employed to cover the whole tuning range of Cr:LiSAF. The first set have a high reflectivity covering the 750–850 nm region, where as the second set provided reflectivity in the 900 to 1050 nm band. Both of the pump mirrors have transmission above 95% at the pump wavelength (660 nm). In the cw tuning experiments, a 4 mm thick off-surface optical axis crystal quartz birefringent filter with a diving angle of 25° is used to tune the laser wavelength [45, 46].

For intracavity frequency-doubling experiments, the cavity is extended via insertion of two more curved highly reflective mirrors (M4 and M5). BBO crystals are placed at this second focus generated by these new curved mirrors. Similar to the pump mirrors, two different sets of curved mirrors are used to cover the full spectral range. The first set has a radius of curvature of 100 mm, and possesses a reflectivity band covering the 750–850 nm region, and a transmission of 95% in the 380–430 nm region (Layertec 105440). The second set has a radius of curvature of 50 mm, a reflectivity band covering 900–1000 nm region, and a transmission band in the 450–500 nm spectral range (Layertec 116786). The intracavity Cr:LiSAF beam size at the focus generated by M4 and M5 is estimated to be around 25 μm and 60 μm for the 50 mm and 100 mm radius of curvature mirrors, respectively.

Since the phase-matching acceptance bandwidth as well as the coating working ranges of BBO crystals are rather narrow, 6 different samples with central design wavelengths of 800, 825, 850, 900, 950 and 1000 nm are used to cover the full-tuning range of Cr:LiSAF. The crystals are optimized for type I phase matching, and the p-polarized intracavity Cr:LiSAF beam is along the ordinary axis of the BBO, whereas the s-polarized second harmonic is an extraordinary beam. Corresponding cut angles (θ) and the absolute value for the effective nonlinear coefficients of the BBO crystals are 29.2° and 1.96 pm/V for 800 nm, 28.4° and 1.97 pm/V for 825 nm, 27.5° and 1.99 pm/V for 850 nm, 26.1° and 2.01 pm/V for 900 nm, 24.9° and 2.03 pm/V for 950 nm and 23.9° and 2.04 pm/V for 1000 nm, respectively. Note that the effective nonlinear coefficient (deff) for type I phase matching in BBO could be calculated using:

$${d}_{eff}={d}_{31}Sin\left(\theta \right)-{d}_{22}Cos\left(\theta \right)Sin\left(3\varphi \right),$$
(1)

where d31 and d22 have values of 0.08 pm/V and 2.2 pm/V, respectively. The value of \(\theta\) is dictated by phase-matching condition, and \(\varphi\) could be chosen freely as 90° to maximize effective nonlinearity, resulting in:

$${d}_{eff}={d}_{22}Cos\left(\theta \right)-{d}_{31}Sin\left(\theta \right).$$
(2)

Note that since the phase-matching angle differs only slightly in the 800–1000 nm range (29.2–23.9°), the effective nonlinearity of the BBO stays almost constant at around 2 pm/V for all the crystals. The BBO crystals used are flat–flat cut and antireflection coated at both the fundamental and second-harmonic wavelengths, and have a length of 2 mm, except the 800 nm crystal which has a thickness of 4 mm. To generate single blue output, a 150 mm radius of curvature metallic high reflector (M6) is employed to retro-reflect back blue power transmitted from M3. To boost up the intracavity laser powers as well as second-harmonic generation efficiency, the OC is replaced by a high reflective mirror (M3). As an alternative, one can still use a very lowly transmitting output coupler (~ 0.1%) to simultaneously obtain near-infrared and visible output form the laser system.

3 Experimental results and discussion

3.1 Continuous-wave lasing and tuning in the near infrared

In this section, we will first present cw lasing performance of Cr:LiSAF in its main operation region (at the fundamental wavelength). Figure 2 shows the measured variation of Cr:LiSAF laser output power as a function of absorbed pump power for six different OCs with transmissions ranging from 0.015 to 3%. The 1 and 3% transmitting output couplers are optimized for lasing around the gain peak (850 nm), whereas the 0.015–0.6% couplers are designed to operate near 1000 nm. The best laser performance is obtained with the 1% transmitting OC. Using this coupler, up to 190 mW of cw output is obtained around 852 nm at an absorbed pump power of 365 mW. Due to the usage of single-mode diode pump sources, the Cr:LiSAF laser output is also single-mode (single transverse mode), with a TEM00 beam profile. The lasing threshold is 16 mW, and the slope efficiency with respect to absorbed pump power is 54%. This slope efficiency is one of the highest efficiencies yet reported from Cr:LiSAF proving the output beam quality of the single-mode laser didoes [4]. For the 0.015–0.6% transmitting output couplers, the free running wavelength is shifted to 1000 nm region (as also indicated in the figure). In this regime, best performance is obtained with the 0.6% transmitting output coupler. Using the 0.6% OC, we have achieved a cw lasing threshold of 40 mW, and a cw output power of 95 mW is acquired with a slope efficiency of 30%. Here, as expected, at this long-wavelength regime, the lasing threshold increases and the lasing slope efficiency decreases due to the shift of the laser wavelength far-off the gain peak [47, 48].

Fig. 2
figure 2

Continuous-wave power efficiency curves for the Cr:LiSAF laser taken at 0.015, 0.15, 0.3, 0.6, 1, and 3% output coupling. Free running laser wavelength is indicated for each case

Full size image

We have estimated the round-trip intracavity laser resonator losses (L) and the intrinsic laser slope efficiency (\({\eta }_{O}\)) at a lasing wavelength of around (∼1000 nm) using Caird analysis approach. According to Caird analysis [49, 50], the slope efficiency η of the laser can be expressed as:

$$\eta =\left[\left(\frac{h{\upsilon }_{l}}{h{\upsilon }_{p}}\right){\eta }_{p}\left(\frac{{\sigma }_{e}-{\sigma }_{esa}}{{\sigma }_{e}}\right)\right]\frac{T}{T+L}={\eta }_{O}\frac{T}{T+L},$$
(3)

where \(h\) is Planck’s constant, \({\upsilon }_{l}\) (\({\upsilon }_{p}\)) is the laser (pump) photon frequency, \({\eta }_{p}\) is pumping efficiency, \({\sigma }_{e}\) is the emission cross section, \({\sigma }_{esa}\) is the excited state absorption cross section, and T is the transmission of the OC. A plot of the experimentally measured variation of the inverse slope efficiency (1∕η) as a function of inverse output coupling (1∕T), hence, gives a straight line (Fig. 3), and L and \({\eta }_{O}\) can be determined in a straightforward way. Using Eq. (3), the best-fit values of L and η0 were determined to be (0.25 ± 0.05) % and (44 ± 5) %, respectively. Note that in this analysis, we have only used lasing data around 1000 nm, hence, the 1 and 3% OC data are excluded. For the 1 cm long crystal, the estimated passive loss is then only around 0.1% per cm, proving the advantage of Cr:LiSAF in constructing high-Q-cavities. Note that the estimated intrinsic slope efficiency (44%) is for 1000 nm operation, and is lower than the ones generally reported (54%) for 850 nm operation [4].

Fig. 3
figure 3

Measured variation of the inverse of the slope efficiency (1∕η) with the inverse of the output coupling percentage (1∕T). Employing Caird analysis, we have estimated the round-trip passive cavity loss (L) and intrinsic slope efficiency (η0) to be (0.25 ± 0.05) % and (44 ± 5) %, respectively

Full size image

Figure 4 shows the measured cw tuning range of Cr:LiSAF laser using OCs with 0.015, 0.15, 0.3, and 1% transmission at ~ 365 mW absorbed pump power. The normalized emission cross-section curve of Cr:LiSAF for E//c is also shown. Relatively smooth tuning of the laser from 795 to 1010 nm is demonstrated and as expected the highest output power is obtained near 850 nm, close to the gain peak of the material. The output of the Cr:LiSAF laser is multi-longitudinal mode. On the other hand, due to the presence of BRF in the cavity, the spectral width of the Cr:LiSAF laser output is narrower than 0.2 nm in the whole tuning range, and is even below 0.05 nm for most wavelengths (measurement limited by the resolution of the spectrometer at hand). The limited reflectivity bandwidth of our pump mirrors prevented lasing below 795 nm (earlier studies showed tuning down to 770 nm [16, 21]). A long-wavelength tuning edge of 1090 nm and 1102 nm is achieved using 0.3 and 0.015% transmitting output couplers, respectively. The reduced emission cross section and hence gain prevents lasing above 1110 nm in Cr:LiSAF. Among the Cr:Colquiriite materials (Cr:LiCAF, Cr:LiSGaF), Cr:LiSAF provides the broadest tuning range in the near infrared, which motivates us to focus on it Cr:LiSAF in this broadly tunable frequency-doubling study [4].

Fig. 4
figure 4

CW tuning curve of the SMD pumped Cr:LiSAF laser at 365 mW absorbed pump power level. The data are taken with four different output couplers with transmissions between 0.015 and 1%. The reflectivity range of each output coupler is listed in the picture legend. The normalized emission cross-section data for E//c-axis of Cr:LiSAF are also shown for comparison

Full size image

3.2 Intracavity second-harmonic tuning results in the visible

Figure 5 summarizes the cw intracavity frequency-doubling performance of Cr:LiSAF laser. The data are taken at an absorbed pump power of 365 mW, using six different BBO crystals with central operation wavelengths of 800, 825, 850, 900, 950 and 1000 nm. For each BBO crystal, the BRF angle is first varied to tune the central wavelength of the Cr:LiSAF laser. Then for each wavelength, the second-harmonic efficiency is optimized by playing with the tilt and position of the BBO crystal. This is required because BBO has a very small angular phase-matching bandwidth [51]. As an example, the 2 mm thick 850 and 1000 nm BBO crystals have an estimated frequency-doubling acceptance bandwidth of only around 7.2 and 16 nm, respectively. Note that operating the nonlinear crystal away from its design wavelength requires usage of the sample away from normal incidence, which increases the BBO crystals losses, and reduces Cr:LiSAF laser intracavity powers (the high-Q laser cavity is extremely sensitive to losses). As a result, the tuning curve for each crystal has a bandwidth of around 10 nm, and beyond that the generated frequency-doubled powers reduce considerably. On the positive side, BBO has quite broad phase-matching temperature bandwidth, and hence active temperature control of the nonlinear crystal is not required (local internal heating of the BBO crystal does not result in a variation of efficiency) [51].

Fig. 5
figure 5

Measured cw intracavity frequency-doubled tuning (402–535 nm) performance of the SMD pumped Cr:LiSAF laser at an absorbed pump power of 365 mW. The data are taken with 6 different BBO crystals with central operation ranges between 800 and 1000 nm. The gray curve is the estimated tuning performance based on cw tuning characteristics of the Cr:LiSAF laser in its fundamental wavelength range

Full size image

As we can see from Fig. 5, an overall second-harmonic tuning range covering the 402–535 nm region is achieved in this work. Best performance is obtained at a wavelength of 422.5 nm, where we have reached up to 17.5 mW of output power. This corresponds to an optical-to-optical conversion efficiency of 4.2% (420 mW total power is considered here, efficiency with respect to absorbed pump power is 4.8%). Since, none of the cavity elements require active temperature control, considering the 33% electrical efficiency of the pump diodes, the overall electrical-to-optical conversion efficiency of the system is also rather high (1.4%).

A conversion efficiency of 16% and an output power of 1.16 W have been reported at 430 nm from a intracavity frequency-doubled Cr:LiSAF laser pumped by four 1.8 W multimode diodes (7.2 W total pump power) [16]. The higher efficiency of the earlier study is due to the availability of 8 times higher pump powers (the efficiency of the nonlinear process scales with the fluence at low conversion regime). Basically, the higher efficiencies obtained in [16] are not achievable in the current compact, low-power diode-pumped system due to the limited intracavity power levels. On the other hand, a rather complex system with diode and crystal cooling was used in [16], and the demonstrated electrical-to-optical conversion efficiency of this work (1.4%) is an order of magnitude above what can be achieved with high power diode-pumped systems (as diodes and crystal require active cooling, and use considerable amount of electricity). Overall, the results obtained in this work show that the state-of-the-art SMD diode-pumped systems have the potential to replace more expensive and bulkier multimode diode-pumped systems in future years.

In terms of the achieved tuning range (402–535 nm), to our knowledge, this study reports the broadest second-harmonic tuning range ever reported from cw Cr:LiSAF lasers, and covers all the spectral regions from violet to green (Fig. 6 shows sample optical spectra taken along the tuning range). A tuning interval between 427 and 443 [38], 432 and 442 nm [39], and 387 and 463 nm [16] was achieved earlier. On the short wavelength side, the tuning limit in our work (402 nm) is restricted by the reflectivity of the optics that were used (a cw tuning limit of 387 nm is reported in [16]). On the long-wavelength side, we could reach wavelengths up to 535 nm. This extends the reported cw intracavity second-harmonic tuning edge of Cr:LiSAF lasers from 463 nm [16] to 535 nm: a 72 nm extension into the cyan and green regions of the spectrum.

Fig. 6
figure 6

Sample optical spectra from the cw intracavity frequency-doubled Cr:LiSAF laser showing tunability of the second-harmonic beam in the 402–535 nm region

Full size image

Note that in Fig. 5, we also show the estimated normalized blue-conversion efficiency as a function of wavelength. The estimation curve is based on the measured cw tuning characteristics of the Cr:LiSAF laser, and assumes that the achievable second-harmonic power levels scales with the square of the available Cr:LiSAF laser power at the fundamental wavelength. As we can see, in general there is a very good agreement between the measured cw second-harmonic powers and the estimated performance. We believe that, the slightly higher efficiency at 475 nm compared to 450 nm is partially due to the better spectral properties of cavity mirrors (M4-M5) in the 950 nm and 475 nm spectral regions (better reflectivity at 950 nm and higher transmission around 475 nm: Layertec mirror 116786). The gaps in the experimental data are due to limited tuning range of BBO samples, and could be improved in future studies. As an example, low efficiency at 440, 465 or 525 nm is due to the lack of BBO crystals with a central wavelength of 880, 930 and 1050 nm during the time of the study. Basically, since Cr:LiSAF laser is very sensitive to increased losses, optimization of the performance at different wavelengths require usage of nonlinear crystals with specifically designed coatings for the region. Nonlinear crystals with broader coating bandwidth could enable larger tuning ranges in future studies. On the other hand, broader coatings usually have larger losses, which might reduce the conversion efficiencies.

4 Summary

In conclusion, we have first investigated cw lasing performance of a compact Cr:LiSAF laser pumped by two low-cost state-of-the-art 210 mW single-mode diodes. We have shown that, this simple system provides cw powers approaching 200 mW, could reach slope efficiencies above 50%, and provides smooth tuning in the 795 to 1103 nm range. Upon frequency-doubling with BBO crystals, the tuning in the near infrared is effectively transferred into the visible region, where output powers up to 17 mW and a tuning range covering the 402–535 nm region are demonstrated. The system has an electrical-to-optical conversion efficiency as high as 15 and 1.4% for the near infrared and visible operation regimes, respectively. This broadly tunable compact and efficient system could ideally be powered by simple AA type batteries. The system has many advantages compared to Ti:Sapphire-based approaches, where much higher pump powers are required due to the intrinsically high passive losses of the crystal [52, 53]. In terms of electrical-to-optical conversion efficiency, other members of Cr:Colquiriites such as Cr:LiCAF could provide a similar performance, but the tuning range of Cr:LiCAF is rather limited (375–433 nm) [30, 54, 55] compared to what we have achieved here with Cr:LiSAF (402–535 nm).