1

– INTRODUCTION

It may look somewhat inadequate to report on a particular optical component in a conference devoted to aerospace techniques and systems. Optical gratings have long been around in spectroscopy, spectral analysis, lasers. The interest in gratings has rapidly developed during the last decade as a consequence of the move of planar manufacturing technologies to shorter feature size, of the need of more compact and lower cost opto-electro-mechanical systems, and also of the discovery of new resonant diffractive effects which provide periodical structures with improved characteristics and new functionalities ⁽¹⁾.

Diffractive gratings as one of the members of the larger family of Diffractive Optical Elements (DOEs) thus represent an enabling technology for system of smaller size, lower weight, less sensitive to temperature and vibration, more reliable and of lower cost. Such features are highly desirable in aerospatial applications. We are here going to illustrate in two very different examples how optical gratings can help solve technical problems which can be met in aerospace. The first one is an example of displacement measurement function integrated into mechanical modules. The second one is an example of how a grating can stabilize the polarization of a microchip laser for inter-satellite communications.

2

– OPTICAL GRATINGS FOR NON-OBSTRUSIVE DISPLACEMENT MEASUREMENT

The usual breakdown of a monitored (electro)mechanical translation or rotation module is composed of three parts: the fixed mechanical part, the moving part, and the measurement system. The latter is in turn composed of at least three parts : the intelligent part comprising a light source, detectors and means to encode the displacement (usually two or three gratings), and two parts mobile relative to each other which will be fixed onto the module to be monitored with well defined standards. This is industrially correct but represents a practically stiff and unwieldy solution. There are applications and markets where the price to pay in term of weight and size is too large. Aerospace is one of them. The alternative solution is to get rid of the supplementary mechanical parts and constraints related which the displacement sensor by itself represents, and to adopt instead a concept whereby the displacement measurement function is inserted non obtrusively into the (electro)mechanical system in the form of a miniaturized optical element without imposing constraints to the designer. One condition for this is the miniaturization of the read head. One further advantageous consequence of the read head being insertable into the core of the (electro)mechanical system or module is that this enables the measurement to be made where the displacement actually takes place, i.e., before elastic strains, backlash and frictions distort the desired information.

Figure one is the illustration of one of the possible embodiments of the basic principle of diffractive interferometry⁽²⁾ whereby the collimated beam of a semiconductor laser crosses the readout grating a first time, then gets diffracted into the +1 and –1 reflected orders by the moving measurement grating. The latter are both diffracted into the 0^th and the –1^st transmitted orders of the readout grating. The spatial period Λ₁ of the readout grating is smaller than the beam wavelength l so that there is no transmitted diffraction order except the 0^th order under normal incidence from the substrate side. If the period Λ₂ of the measurement grating is exactly half that of the readout grating, the latter acts as a recombining element projecting the beams back-diffracted by the measurement grating onto each other in two directions where they interfere. Two detectors placed at the back side of the readout grating substrate at a distance where the two interfered beams have separated spatially detect the interference products. The interference products depend on the phase between the two interfering orders emanating from the measurement grating. If the incidence is normal, the phaseshift only comes from the lateral displacement of the measurement grating with respect to the read head, or conversely. It is knows that the phase-shift δφ experienced by a wave diffracted in the order m of a moving grating of period Λ over the displacement Δx is Δφ = mk_g Δx where k_g = 2π/Λ is the grating spatial frequency. Refering now to figure 1, the phase-shift Δφ between the two beams of order + and – 1 diffracted by the measurement grating having moved by Δx is Δφ = 2k_g2 Δx where k_g2 = 2π/Λ₂. The optical power detected by the detectors is composed of a DC term plus the interesting Δφ-dependent AC term. The electrical period of the AC term is exactly half that of the measurement grating, Λ₂/2. The readout grating can be advantageously composed of two sub-gratings of identical period Λ, but displaced by Λ₁/4 relatively to each other. Four detectors suitably placed around the grating at the backside of its substrate provide four sinusoidal electric signals of Δφ in quadrature which allows displacement direction detection and fine interpolation. Taking as an example Λ₂ = 1µm, Λ₁ = 0.5µm (the wavelength can preferably be between 600 and 800 nm) a displacement resolution of 125 nm is obtained by just counting the zero crossings of the sine and cosine signals.

Figure 1 :

Perspective view of a miniaturizable translation sensor with oblique incidence using two gratings in a period ratio of 2

Figure 2 is the picture of a particular embodiment of the described principle on the way towards miniaturization. Its 2.5 cm³ volume contains a laser diode with oblique incidence, the readout grating, four detectors and elementary signal conditioning. The module is designed so as to incorporate an Opto Asic performing the photodetection, the signal processing and fringe counting as well as interpolation. The OptoASIC will be described in a later paper; early experimental characteristics will be presented at the conference. For proper operation and exact displacement measurement the two gratings must be properly adjusted. The measurement is independent of the grating spacing but a strict parallelism must be ensured between them. Roulis is not a matter of concern and the effects of pitch are usually kept under control by a conventional translation mechanics. The most critical displacement characteristics is the yaw which causes a cosinus error but more importantly leads to a fading of the interference signal picked up by the detector due to the formation of fringes on its surface. Typically, the parallelism between the lines of the two gratings must be kept within a few minutes of arc. Interestingly, the tolerance increases with the miniaturization of the read head.

Figure 2 :

Encapsulated read head of the type depicted in figure 1 containing a laser, the readout grating and an opto ASIC signal processing

Figure 3 illustrates the measured signal contrast versus the angle between grating lines. The miniaturization potential of this detection principle is high depending on the type of implementation. A LED can be used instead of a laser⁽³⁾ and an OptoASIC can dramatically reduce the number of parts and the weight of a sensor system. As soon as the main parts of the system (the readout and measurement gratings) can be cheaply and reproducibly produced in the form of small items, this measurement technique can be given a large number of diversified applications where the measurement function must be performed non-obstrusively where the displacement actually takes place.

Figure 3 :

Dependence of the interference contrast on the tilt angle between the lines of the two parallel gratings

3-

A GRATING STABILIZES THE POLARIZATION OF MICROCHIP LASERS

A second, completely different, application example shows how a grating can monolithically perform a polarizing function on the emission of an otherwise arbitrarily polarized laser. Microchip YAG lasers for instance appear to be light sources exhibiting all together a number of practically very interesting characteristics and are bound to satisfy the needs of a very diverse and large market⁽⁴⁾. One important function can so far not be performed in a monolithic form. It is the polarization control of the emission. Polarization must be stabilized in many applications such as interferometry, nonlinear effects such as second harmonic generation, parametric oscillation, and coherent communications just to quote a few. We show here that an extra-cavity waveguide grating added to the standard multilayer laser coupler can under a proper design lead to a sufficiently strong difference of reflection coefficient between the selected and undesired polarization⁽⁵⁾. This can advantageously replace the cumbersome stress induced birefringence exerted externally onto the laser crystal. Although the direct advantage of such solution for aerospace is not direct, it illustrates how new types of gratings can perform new functions which their classical countersparts can not.

Its principle relies upon a waveguide grating interference mechanism whereby the reflection of an incident wave at the corrugated surface can be set to between 0 and 100 % by a suitable choice of the structure parameters. The device comprises a Nd : YAG chip, a standard multilayer mirror, a last pair of layers, and a corrugation grating in the last high index layer at the air side as illustrated in figure 4.

Figure 4 :

Multilayer mirror with a polarising corrugation grating in the last high index layer.

The last pair of low and high index layers plays an important role in the polarizing mechanism. It is thicker than the other pairs of layers of the standard mirror. Its optical thickness of λ does not disturb the reflection characteristics which the standard multidielectric mirror exhibits for both incident polarisations. As shown in Figure 5, the grating period is so designed that the undesired TE polarization is coupled into a waveguide mode of the last high index layer and re– coupled into the cavity with a phaseshift of π.

Figure 5 :

How a suitably designed grating can induce differential reflection in a multilayer mirror

The grating does not couple the TM polarization into a TM mode of the high index layer because of phase mismatch. The TM polarization therefore does not « see » the grating and experiences the same reflection as if the grating was absent. It does not even diffract into the high index YAG substrate in the form of free space diffraction orders. With a multilayer composed of Ta₂O₅ and SiO₂, the grating period was set to 620 nm and the groove depth to 140 nm (the low index layer has 1.48 index and 180 nm thickness, the high index layer has 2.18 index and 120 nm thickness).

Since the polarizing multilayer mirror is associated with the active medium of the laser cavity, the reflection coefficient differential does not have to be large : a 20 % damping of the reflection coefficient of the undesired polarization is sufficient to prevent its lasing.

Such grating mirror was fabricated by means of direct e-beam writing and RIBE–etched into Ta₂O₅ (Friedrich-Schiller-Universität, Jena) and tested as the output coupler of a microchip Nd : YAG laser. Early results show that the ratio between the TE and TM polarisations is below –17 dB. This demonstrates that a grating can be the integrated element which stabilises the polarisation state of a solid state microchip laser. This implies that manufacturing of polarisation stabilised microchip lasers can be fully achieved by low cost planar technologies. This technique can be applied to pulsed as well as to CW lasers.

4

-CONCLUSION

Optical gratings, and to a larger extent Diffractive Optical Elements represent for aerospace applications an enabling technology allowing lower weight, more compact and more reliable optical components, modules and instruments. Two examples of optical grating applications have illustrated the potential of diffractive optical elements for aerospace : the first one deals with displacement sensing where it is shown that very small size and light weight read heads can track the relative movement between mechanical parts where the later takes place and without the need of bulky and heavy interface parts. The second one illustrates how a monolithic diffraction grating integrated to a light source can improve its emission characteristics with more reliability than the state of the art.

Beyond these two examples much more could be said about the potential of more general DOEs with higher level functionality for aerospace, the key words being small size, low weight and high reliability.