The present invention relates to coherently combined lasers and in particular, though not exclusively, to coherent beam combination in an array of photonic crystal surface emitting lasers ( PCSELS ) surface-emitting diode lasers using nearest neighbour phase measurement and phase control .

By the technique known as Coherent Beam Combination, arrays of mutually-coherent laser beams can be spatially tiled to provide a beam that has higher power and higher brightness than an individual beam and is steerable . Such arrays can also provide beams that are focused on a distant target , or a beam whose far- field beam profile is electronically configurable .

In order to achieve high brightness , beam steering or control of the far field, the relative phases of the individual beams must be precisely controlled . Typically, for maximum brightness , the individual beams , known as the subaperture beams , must be co-phased . To steer the beam, a phase tilt ( linear variation of phase with transverse position) is applied across the entire beam . To shape the beam, an additional per- subaperture phase deviation is applied in order to generate a far- field of a speci fic shape . The far- field can be trans formed into a near- field at a process plane using a focussing lens - often called a Fourier lens .

' Sel f-Synchronous and Sel f-Referenced Coherent Beam Combination for Large Optical Arrays ' , Thomas M . Shay et al , IEEE JOURNAL OF SELECTED TOPICS IN QUANTUM ELECTRONICS , VOL . 13 , NO . 3 , MAY/ JUNE 2007 describes a homodyne technique using a single detector to phase lock multiple laser fibre sources without requiring a reference beam referred to as LOCSET (locking of optical coherence by single-detector electronic- frequency tagging) .

Each subaperture beam emits at a discreet frequency which limits the number of subaperture beams that can be locked together .

' Towards Ultimate High-Power Scaling : Coherent Beam Combining of Fiber Lasers ' , Hossein Fathi et al , Photonics 2021 , 8 , 566 includes a description of the Stochastic Parallel Gradient Descent ( SPGD) method, being a homodyne method of detection using a single detector measuring the far field profile of the combined source . An SPGD based optimisation algorithm measures the brightness of the combined source and perturbates the phase of each one to increase the brightness . The algorithm runs until a maximum brightness is achieved . Increasing the number of subaperture beams , increases the number of potential perturbations until maximum brightness is achieved, limiting the maximum number of subaperture beams that can be phase locked in a quick and ef ficient manner .

'Coherent Beam Combining Using an Internally Sensed Optical Phased Array of Frequency-Of fset Phase Locked Lasers ' , Lyle E . Roberts et al , Photonics 2020 , 7 , 118 discusses optical heterodyne detection using a frequency shi fted reference beam as a master oscillator which all the other channels are locked to . Since all subaperture beams are locked to a single master there is no physical limit on the scalability of this technique . However, the use of an independent reference source along with the need for a detector per subaperture beam adds complexity, si ze and cost to the system .

'Collective phase measurement of an array of fiber lasers by quadriwave lateral shearing interferometry for coherent beam combining' , C . Bellanger et al , OPTICS LETTERS , Vol . 35 , No . 23 , p3931 , December 1 , 2010 , describes a quadriwave lateral shearing interferometer for phase detection in beam-combining fibres disposed in a matrix arrangement . The phase step between adj acent fibres in two dimensions is accessible to provide fast treatment by a spatial demodulation process , and the phase map from the fibres can be estimated in real time . No external reference is required . These sel f-locking fibre laser arrays are limited in their scalability as each fibre laser has a Fabry- Perot ( FP ) cavity with close positioning of linewidths . For sel f-locking the lasing medium gain spectra and the FP comb spectra of each cavity must align so that they all lase on the same line . This has been found to be unfeasible for more than ten fibre lasers .

WO 2020/ 016824 discloses a coherent beam combination ( CBC ) system including an array of fibre laser beam sources generating coherent beams directed towards a target . The phase modulators allow adj ustment of relative phase of fsets of the beams . A detector monitors an intensity of the radiation impinging on an area of the target . A controller receives the intensity parameter and controls a phase adj ustment of the beams according to a deterministic ( i . e . , quantitative ) measurement of a phase of fset of each beam relative to a representative phase of the sum of all the other beams . This is achieved by using interferometric techniques , referred to as Target In-the-Loop Interferometry ( TILI ) . Bulky optics are required to operate this arrangement .

WO 2016/ 027105 discloses a laser structure comprising a first photonic crystal surface emitting laser ( PCSEL ) , a second PCSEL, and a coupling region that extends between the first PCSEL and the second PCSEL along a longitudinal axis and that is electrically controllable so as to be capable of coherently coupling the first PCSEL to the second PCSEL . Each PCSEL include an active layer, a photonic crystal , and a two-dimensional periodic array distributed in an array plane parallel to the longitudinal axis within the photonic crystal where the two- dimensional periodic array is formed of regions having a refractive index that is di f ferent to the surrounding photonic crystal . This process is limited to two PCSEL .

It is an obj ect of the present invention to provide a method and apparatus of coherent beam combination which obviates or mitigates at least some of the disadvantages of the prior art .

According to a first aspect of the present invention there is provided a method of coherent beam combination comprising the steps :

( a ) providing a two-dimensional monolithic array of laser sources , the laser sources being photonic crystal surface emitting lasers ( PCSEL ) ;

(b ) coupling each laser source to one or more adj acent laser sources to form a sel f-locked laser array with each laser source emitting at the same frequency, wherein the coupling is by waveguides and each waveguide incorporates a phase modulator ;

( c ) measuring a phase di f ference between adj acent laser sources ;

( d) using a phase control algorithm on the measured phase di f ferences to provide control signals to drive the phase modulators ;

( e ) allowing beams from each laser source to propagate in a medium, forming a coherently combined laser array; and

( f ) controlling the relative phase of each laser source to control pointing and wavefront of the coherently- combined laser array .

By using PCSELs , a large number of emitters in a 2D array on a single wafer is achievable making them easier to produce and arrange than for laser fibre and diode sources . They also do not suf fer from the FP frequency comb while providing an extremely narrow linewidth with laser spectra that are very close together, due to it being easier to maintain uni form temperature across a single wafer than between physically separate fibre lasers .

By only making measurements with respect to an adj acent laser source , ef fectively the nearest neighbour, nearest neighbour phase measurement and phase control is realised . A complex high- frequency heterodyne signal detection is not required and the array can be easily scaled .

Preferably, at step ( c ) the phase di f ference between adj acent emitters is measured by a passive phase-diversity technique . In this way, a homodyne technique is used for phase measurement .

Preferably, the phase of each laser source is locked to one or more other laser sources in a sel f-locked array . More preferably, the laser sources are locked to multiple mutually- coupled laser sources in a mesh .

Preferably, the method includes the steps of measuring intensity of each subaperture beam from each laser source and controlling the power to each laser source . In this way, a feedback loop control system is provided to give the coherently combined beam a desired intensity profile .

According to a second aspect of the present invention there is provided a coherent beam combination apparatus comprising : a two-dimensional monolithic array of laser sources , the laser sources being photonic crystal surface emitting lasers ( PCSEL ) providing an array of subaperture beams directed towards a target ; a plurality of coupling systems being waveguides arranged between adj acent laser sources with each laser source being coupled to one or more adj acent laser sources ; each optical incorporates a phase modulator ; a phase detector to measure a picked-of f portion from the subaperture beams ; a medium through which beams from the laser sources can propagate ; and a phase control device including a processor to drive the phase modulators and adj ust the relative phases of the laser sources to control pointing and wavefront of the coherently combined laser array .

In this way, an array of subaperture beams can form a coherently- combined beam by using nearest neighbour phase measurement and phase control between pairs of adj acent PCSELs ( Photonic Crystal Surface Emitting Lasers ) .

Preferably, each optical coupling system is arranged hori zontally or vertically in the array in a regular grid pattern or a hexagonal pattern . In this way, all the optical coupling systems may be identical . Preferably each laser source is optically coupled to all its adj acent laser sources . In this way, laser sources within the array will have four couplings while those on the edges will have three and the corners have two . Alternatively, each laser source is coupled to a maximum of two others so that cascaded pathways are created over the array . This reduces the amount of phase modulators required .

Preferably, the detector is a phase detector . The detector may also be an intensity and/or power detector . The apparatus may include a partially reflecting mirror to provide the picked-of f portion of the subaperture beams . The apparatus may also include a power control unit including a processor connected to a laser drive unit for adj usting the power to each laser source . In this way the intensity profile of the coherently combined beam can be controlled .

The apparatus may include one or more optical elements to focus the coherently-combined beam onto a target such as a workpiece . The one or more optical elements may comprise a telescope . Preferably the one or more optical elements comprise a focussing lens which may be considered as a Fourier lens .

Embodiments of the present invention will now be described, by way of example only, with reference to the following drawings of which :

Figure 1 is a schematic illustration of a coherent beam combination apparatus according to an embodiment of the present invention;

Figure 2 is a graphic illustration of coupling between laser sources for in a 5x5 array according to an embodiment of the present invention;

Figure 3 is a graphic illustration of aperture overlap for passive phase-stepped interferometry according to an embodiment of the present invention;

Figure 4 is an illustration of a phase measurement geometry for a 5x5 array of subaperture beams according to a further embodiment of the present invention;

Figure 5 is an il lustration of a 5x5 array of subaperture beams according to a further embodiment of the present invention; and Figure 6 is an illustration of the phase measurement geometry for the 5x5 array of subaperture beams of Figure 5 .

Reference is initially made to Figure 1 of the drawings which illustrates a coherent beam combination apparatus , generally indicated by reference numeral 10 , according to an embodiment of the present invention . Apparatus 10 includes an array of laser sources 12a-d, with each laser source 12a-d providing a subaperture beam 14a-d, which provide a coherently combined beam 16 directed towards a target 18 which may be a workpiece as is known in the art . The laser sources 12a-d are PCSELs ( Photonic Crystal Surface Emitting Laser ) . The laser sources 12a-d are shown in a linear array of four sources , but it will be appreciated that this is for illustrative purposes and the array is two dimensional , with much greater number of laser sources possibly up to 1000 . Each laser source 12a, d is optically coupled to its nearest neighbour i . e . 12a is coupled to 12b ; 12b is coupled to 12a and 12c ; 12c is coupled to 12b and 12d; and 12d is optically coupled to 12c . The optical coupling system 20a-c between neighbouring or adj acent laser sources 12a-d is a waveguide . For the embodiment shown, a wafer of PCSELs is made where pairs of adj acent PCSELs are interconnected with a waveguide section .

Referring to Figure 2 there is illustrated a two dimensional array of 25 laser sources with the coupling illustrated between every adj acent laser source by a double arrow . While a regular array is shown, those skilled in the art will appreciate that hexagonal patterns of laser sources can be formed on a wafer and thus these alternative patterns can be used so long as coupling exists in at least one path between laser sources .

Each optical coupling system 20a, c can be individually phase modulated 22a-c from a phase modulator driver 24 . As is known in the art , a laser drive unit 26 is used to power the laser sources 12a-d so that they lase . Cooling can be provided to the laser sources 12a-d to remove heat and keep the temperature within an acceptable range . This will achieve temperature uni formity across the laser sources and can be done passively or actively .

In the path of the subaperture beams 14a-d is a partially reflecting mirror 28 arranged at 45 degrees . This mirror 28 will 'pick-of f ' a portion of each of the beams 14a-d and an optical element 30 such as a di f fraction grating, a lens array or pyramid array is used to create overlapping beams onto an array detector 32 being a focal plane array camera . Detector 32 could be a CMOS ( Complementary Metal Oxide Semiconductor ) or CCD ( Charge Coupled Device ) sensor .

A phase detection unit 34 , measures the phase di f ference between adj acent pairs of beams 12a-d using a passive homodyne technique , by generating a static interferogram from the picked of f part of each beam in the pair, sampling the intensity of this interferogram in a phase-diverse way, either with four samples each separated by roughly of pi/2 or with a larger number of samples providing a robust means of phase recovery . The useful interferograms are shown in the shaded regions of Figure 3 . A more compact geometry can be achieved by using back-reflection from downstream optics to generate interferograms in the same configuration . In this case , the sensor array requires clear aperture to pass the transmitted beams (white areas in Figure 3 ) and detector elements to capture the interferograms ( e . g . striped pin or CMOS detectors ) in the shaded regions .

A description of a passive phase-stepping technique for an array of 5x5 subapertures 14 will now be given with reference to Figure 2 . A double arrow represents coupling between the laser sources 12a, b . Here it is seen that every laser source is paired to all available adj acent , nearest neighbour, laser sources in both the hori zontal and vertical directions . The central sources have four connections , those at the edges three couplings and the corners have two . This approach is analogous to the Hudgin geometry for wavefront measurement , in which a continuous wavefront is estimated from phase di f ferences between adj acent sampled points on the wavefront . Methods for reconstructing the phase at the nodal points based on matrix trans formation are well-known and well-characterised .

Figure 4 describes the measured phases in nearest neighbour wavefront measurement , as used in Hudgin wavefront reconstruction . The signals S , S^, S% are considered as phase or wavefront slopes in wavefront sensing but in this invention they represent phase di f ferences between adj acent subapertures 14 . In this invention, instead of reconstructing a complete wavefront , the phase for each subaperture 14 is calculated, relative to some arbitrary common reference phase . The same algorithm as for Hudgin wavefront reconstruction can be used to calculate this , since Hudgin wavefront reconstruction essentially calculates the wavefront at the centres of the nodes . The phases of all subapertures 14 can therefore be calculated via a matrix trans formation .

Any element can be set arbitrarily to zero , or the average can be set to zero . The vector of measured subaperture phases is calculated through wavefront reconstruction, typically by matrix trans formation W the vector of all phase di f ferences S . The vector of phase error signals <5 _e is then calculated by subtracting <5 _m from vector of setpoint ( goal ) phase values .

For a t x t array, Hudgin wavefront reconstruction gives a phase error that depends very weakly on t :

A processor 36 uses a wavefront reconstruction algorithm to recover the phase of each laser source 12a-d relative to a common reference . This reference phase may be notional , it need not be the phase of any speci fic subaperture . This reference could be a speci fic PCSEL, or the average phase of all PCSELs , or any other value , since it is only the variation on phase across the ensemble of all lasers that matters . This may be done within a phase control system 38 which in turn is connected to the phase modulator driver unit 24 .

The control signal may be calculated as follows : the relationship between the vector of phase modulation signals and the vector of phase di f ferences is determined by applying an incremental step input ( i . e . a step input resulting in a small fraction of a radian phase change , for example 0 . 1 rad or 0 . 01 rad or 0 . 01 rad) to an individual phase modulator 22a-c in turn and measuring the ef fect on the phase of each of the laser sources 12a-d, relative to the reference phase . This is repeated by applying an incremental step input to each phase modulator 22a-c in the array in turn . The resultant data set is a matrix describing the response of the wavefront to a set of incremental phase perturbations applied via the phase modulators . Inverting this matrix provides the incremental modulation required to generate an incremental change in the wavefront . The measured wavefront is compared to a desired reference wavefront and the di f ference used to generate a wavefront error signal . The matrix trans formation is applied to this wavefront error signal and combined with a set of integrators , according to well-known control system practise . An electronic or software control loop is thus created that converts the set of measured phases into a set of control signals for the phase modulators 22a-c that achieve the desired setpoint wavefront with zero steady-state error .

A setpoint phase is defined for each laser source 12a-d and use the control system 38 to lock each laser source 12a-d to that setpoint phase . Each laser source 12a-d can have either the same setpoint phase or a di f ferent setpoint phase .

Preferably the power, as intensity at the detector 32 , is also measured from each laser source 12a-d and the information used to control , via power control unit 40 , the power level of each laser source 12a-d from the laser drive 26 via a feedback loop, so as to give the overall beam 16 the ideal intensity profile .

The apparatus may also have an optical arrangement 42 to direct the beam 16 to the target 18 workpiece . This may be a telescope of a Fourier focussing lens as is known in the art .

One can make the setpoint phases all the same to generate a coherently-combined aggregate beam 16 with maximum on-axis brightness . Adj usting the setpoint phases can be made to : vary linearly across the array will adj ust the pointing of the aggregate coherently-combined beam 16 ; compensate for phase errors caused by transmission through the atmosphere between the sources 12a-d and the target 18 workpiece ; compensate for phase errors caused by transmission through an optical system 42 with aberrations between the sources 12a-d and the target 18 workpiece ; and, compensate for phase errors caused by transmission through the air or assist gas between the sources 12a-d and the target 18 workpiece .

This method of phase measurement by use of nearest neighbour measurements and modulation of the optical coupling between adj acent laser sources provides a control system architecture that facilitates high-bandwidth control of subaperture beam phases .

In the prior art , when the subaperture beams are generated in a MOPA configuration the phase of each subaperture can be controlled independently . The measured phase for each subaperture is compared to a reference phase , and the di f ference is fed into a closed loop controller . Zero steady-state error can be achieved by including an integral term in the control loop for each subaperture . In practise , phase will be measured at discrete time intervals and control will be done by a discrete time digital control system .

Whereas Figure 4 demonstrated phase control in full mesh sel flocked configuration, where there are multiple coupling paths between any pair of oscillators . Referring now to Figure 5 , in this representation, where adj acent sources 12 have no connecting arrow, there is no coupling . In this example , there is a unique coupl ing path from any one oscillator to any other, either directly between adj acent oscillators or via intermediate oscillators . Many alternative such configurations exist .

In this configuration, the relative phase of all emitters could be measured by measuring only the phase di f ferences between adj acent emitters where couplings exist . In this case , no matrix trans formation is required as simple addition provides all phases relative to that of the master oscillator emitter . Alternatively, any other configuration of nearest-neighbour phase measurement could be used to achieve the same - the phase measurements do not need to coincide with the phase coupling.

A better estimate of phase will, however, be achieved by using the full set of phase differences i.e. 40 phase differences for a 5x5 array. This is the preferred configuration.

So the preferred configuration is:

Calculate 4> _e as above. Then, for each of the control elements shown in Figure 5, calculate a slope error signal vector E = {Ef,E ₂ ,...} et c. and pass this through a control system loop filter to generate the control signal vector 'P^, ... } . The simplest loop filter is an integrator element transforming each element of E to the corresponding element of P i.e. use an integrating loop filter to transform Ef to 'P , E ₂ to etc. Alternative loop filters may be used. Also, a matrix transformation whole elements are filter functions may be used instead of a simple 1-1 transformation.

Alternatively a simplified control system can be applied by driving each individual phase actuator 'P via a loop filter from the corresponding element of S. This approach minimises the amount of computation for each cycle of the control system, at the expense of a poorer estimate of phase difference and so increased noise in the control system, and potentially also reduced control bandwidth.

A principle advantage of the present invention is that it provides a method of coherent beam combination in an array of self-locked PCSEL sources using nearest neighbour phase measurement and phase control. A further advantage of the present invention is that it provides a method of coherent beam combination in an array of laser sources which can be scaled to a high number of laser sources . A still further advantage of the present invention is that it provides a method of coherent beam combination in an array of laser sources which does not require use of complex heterodyne phase measurement techniques .