KLASTECH takes a whole new approach to the problems associated with increasing the second harmonic conversion efficiencies for Diode Pumped Solid State (DPSS) lasers.
In a traditional architecture the non-linear crystal which is the component that converts the frequency, is placed inside the same cavity as the laser gain medium which generates the fundamental wavelength. This results in a natural mismatch between the requirements to generate maximum fundamental wavelength light for conversion whilst at the same time maximising the second harmonic conversion efficiency in order to generate the required wavelength of light. This can be thought of as an impedance mismatch and is particularly acute since the second harmonic generating (SHG) crystal is non-linear in its response to the flux density of the light impinging on it and indeed requires that a maximum value of fundamental wavelength light power be encountered in order for SHG to occur in an efficient way.
Rather than try and solve the symptoms of this mismatch: reduced conversion efficiency, green noise etc. which often require expensive crystals, optics and other classical ‘work arounds´, KLASTECH instead has designed a new architecture that removes the problem in the first place. In this cavity impedance matched layout, the SHG crystal is placed in its own cavity linked to the first through an intra-cavity coupler (see Technology page). Here the fundamental frequency beams circulated in both cavities are offset minutely allowing a piezoelectric controlled mirror in the second cavity to lock at a point whereby optimal constructive intra-cavity coupling can occur between the fundamental frequency standing waves in both cavities. This results in a significantly increased power level at the fundamental wavelength circulating in the second cavity (containing the non-linear crystal) and a corresponding quadratic increase in the production of the second harmonic.
Practically this allows for reduced complexity in design, removal of the ‘green noise´ problem, significantly reduced pump diode power requirements for the same desired output power, much lower heat dissipation and of course significantly reduced cost of manufacture.
When one considers that this technology has resulted in the world´s first and only CW ruby laser by DPSS technology it becomes clear that this technology is truly revolutionary.
The second cavity that contains a frequency doubling crystal is seen from the first cavity that holds an active laser crystal as a resonant reflector. This arrangement is known, when without a non-linear crystal, as a Fox-Smith interferometer. Hence, locking the two above cavities at one of the resonances of the Fox-Smith interferometer for optimal intra-cavity coupling and highest second harmonic generation efficiency, inherently leads to the lasing at only that resonant frequency. This single frequency (single longitudinal mode – SLM) lasing remains as long as the two above cavities remain locked, hence, no mode hopes occur during the stable operation of the laser.
The actual spectral position of a lasing SLM is defined by the frequency of the second cavity resonance at which the two cavities are locked. As the nature of this resonance is a constructive multiple beam interference, its spectral position is sensitive to the change of the optical path difference between the interfering beams due to thermal variation of this dimension in the relevant opto-mechanical components. At standard operational conditions when the lab temperature can vary easily by a few degrees Celsius, the laser cavity components are still kept at a fixed temperature with precision of around 0.10° Celsius. However, even such precise temperature control does not eliminate completely any drift of the spectral position of the lasing longitudinal mode. Depending on the laser emission wavelength, typically this drift is within a range from 1 to 2 pm (from 0.5 to 1.0 GHz). These values can be reduced to around 0.1GHz or lower if the “environmental” temperature, for instance, the temperature of a heatsink where the laser head is placed, is “pre-stabilised”.
For some exotic applications, where the spectral drift of the lasing mode is required to be far below the above figures, our technology allows for a reduction in this spectral drift to be achieved in the range of less 1MHz by means of external locking of the lasing mode to a spectral reference like, for example, a narrow absorption feature of a low pressure gas cell. Such customisation however, has to be discussed in detail for every specific application.
When measuring and understanding single frequency line width at values in the range less than 1MHz, one has to link the issue with a characteristic “exposure” time in a set-up where a single frequency laser is to be used. For instance, if such a laser is used for recording a hologram of a large scale object requiring exposure time up to minutes, the spectral width of the laser emission would appear as comprising of two parts; the first being a short-time (“momentum”) spectrum of the emission and the second being a broadening of this “momentum” spectrum due to physical and technical “instabilities” always present to a certain extent in any set-up. Two major types of instabilities that potentially affect the result of the spectral width measurement are the characteristic transient processes inside the laser active medium and laser cavity, and the external influences on the laser set-up. While the internal transient processes are rather fast and would be averaged at a time scale in the µsec-msec range, showing a short time spectrum width in 10´s kHz -1MHz range, the above thermal frequency drift is an example of a slow instability that would result in an appearance of the laser spectrum at a much higher value if the measurement time is extended. Historically, subject to an application, the line width of the laser emission used to be presented in different units: nm, pm, Hz, wave numbers (1/cm). In our data sheets we present the SLM emission line width in MHz´s. The emission of 1MHz line width corresponds 3.33 x10-5 wave numbers or ca 0.001pm at 532nm wavelength.
DPSS lasers unlike diode lasers are very susceptible to variations in the pump diode power. This is due to a number of reasons but two are particularly critical:
1/ The laser gain medium and the second harmonic conversion crystal are both optically non-linear in response to their respective excitation wavelengths, therefore alterations in the original pump diode output power can have unexpected effects including unstable performance and increased optical noise;
2/ A DPSS laser´s thermo-mechanical stability is controlled very carefully to ensure stable optical performance, varying the pump diode power can affect this thermal equilibrium to the detriment of the laser performance and lifetime.
KLASTECH does not take this approach. Instead we offer two routes to modulation, one is better defined as customer output power adjustment, where we make use of the second cavity to allow the implementation of a simple optical attenuator after the generation of the second harmonic, thus leaving all the thermal and non-linear optical components unchanged.
The other makes use of the fact that the system inherently locks and tunes the second harmonic generating cavity to the first as describe above. Varying the position of this ‘locking´ can result in very good modulation properties again with no change in the system´s thermal equilibrium.
However, as with all DPSS lasers the modulation speed is limited and for high frequency modulation we recommend the implementation of an acousto-optic modulator.
KLASTECH lasers produce second harmonic conversion efficiencies anywhere between 4 and 10 times greater than other intra-cavity architectures. Generally as the overall power of the laser increases the efficiency gains of the KLASTECH technology are reduced.
That said we are confident that we can achieve up to 10W CW in the 532nm wavelength range from our compact DPSS lasers without recourse to any external cooling purely due to the exceptional efficiency of our lasers.
Absolutely. We do not use quasi CW, all our lasers are truly CW in performance.
Describing a laser beam divergence as diffraction limited simply states that the performance of our lasers is limited not by design, but by principal physical restrictions. Laser beams produced by KLASTECH´s DPSS lasers are purely Gaussian fundamental TEMoo mode beams with M˛ <1.05. For such beams the diffraction limited divergence means that the product “the beam divergence measured in far field region multiplied by the Gaussian beam diameter measured at its waist position” holds as an invariant: ΔΩ.ω = 4λ/π, where ΔΩ and ω are the far field beam divergence and the beam waist diameter both measured at the e-2 level of the beam intensity cross-sectional distribution, and λis the laser emission wavelength. When in the above formula λis presented in microns, ω – in mm, ΔΩ is given in mrad. Thus to achieve or to modify the laser beam divergence to a required mrad value one has to implement a telescope to modify the laser beam waist accordingly. In most of our lasers, the output beam diameter is kept close to 1mm as a standard, hence resulting in divergence of around 1mrad, depending on the wavelength. One remark to be added here: the specified divergence is assumed for far field measurement, as for the above formula. For this reason, while measuring divergence of Gaussian laser beams with diameters in the mm range, one has to take into account that the far field region begins at distances 5 - 10m apart from the laser head output aperture. If a divergence measurement is undertaken at shorter distances, in the near field region, also referred to as the Rayleigh range, the result of the measurement will not be consistent and will be dependent on the actual position of the measuring device with respect to the position of the laser beam waist, which is in our lasers, “pinned” to one of the laser cavity mirror surfaces inside the laser head.
We are committed to continuously advancing our laser products through pioneering development and original design.
Christopher J. Madin, CEO