Method and system for forming birefringent voxels

Abstract

A method for forming birefringent voxels includes simultaneously generating a first seed pulse and a first data pulse. The first seed pulse and the first data pulse are spatially separated laser pulses having different amplitudes. The first seed pulse is focused at a first seed location and the data pulse is focused at a first data location. The first seed location and the first data location are separated by a predetermined distance along the scan path, wherein the first seed location precedes the first data location. Subsequently, a second seed pulse and a second data pulse are generated and focused at a second seed position and a second data position, respectively. The second seed is separated from the data location by a predetermined distance. The second data location is the same as the first seed location, resulting in the formation of birefringent voxels.

Description

BACKGROUND

[0001] There is a large demand for data storage. It is estimated that in the near future, cloud storage providers will need data storage capacities in the zettabyte order of magnitude, a zettabyte being one trillion gigabytes (10 21 Most of the data will need to be stored for extended periods of time.

[0002] Examples of currently widely used data storage technologies include hard disk drives, magnetic tape, flash memory, and optical disks. All of these technologies have the disadvantage that data needs to be periodically copied to replacement media. This is costly in terms of energy usage and hardware requirements.

[0003] Magnetic storage media such as hard disk drives and magnetic tape gradually demagnetize. Flash memory is subject to read disturb effects, where repeated reads from a particular flash memory cell cause failure of surrounding flash memory cells. Reflective materials used for data storage in optical media such as DVDs degrade over time.

[0004] Birefringent optical data storage media have been proposed as a solution to these drawbacks. Birefringent optical data storage media comprise a transparent substrate, such as a quartz glass substrate. Data is encoded in three-dimensional nanostructures formed in the substrate. These nanostructures are referred to as voxels.

[0005] A voxel has optical properties that are different from the optical properties of the surrounding bulk substrate. In particular, a voxel is birefringent, in other words, it shows different refractive indices depending on the polarization and / or direction of the incident light. The optical properties of a voxel can be controlled when writing the voxel into the substrate. The optical properties of a voxel are used to encode data.

[0006] Birefringent optical data storage media and their fabrication have been described in, for example, Anderson et al., Glass: A New Media for a New Era? 10th USENIX Conference on File and Storage Technologies (HotStorage 18), 2018, and US 10,236,027 B1. SUMMARY

[0007] In one aspect, a method of forming a birefringent voxel in a transparent substrate is provided. The method comprises: generating simultaneously a first seed pulse and a first data pulse, the first seed pulse and the first data pulse being spatially separated laser pulses, and the amplitude of the first seed pulse being different from the amplitude of the first data pulse; focusing a first seed pulse at a first seed location in the substrate and focusing a first data pulse at a first data location in the substrate, the first seed location being separated from the first data location by a predetermined distance along a scan path, wherein in the scan path the first seed location precedes the first data location; generating a second seed pulse and a second data pulse simultaneously after focusing the first seed pulse at the first seed location and the data pulse at the first data location, the second seed pulse and the second data pulse being spatially separated laser pulses and the amplitude of the second seed pulse being different from the amplitude of the second data pulse; and focusing the second seed pulse at a second seed location in the substrate and focusing the data pulse at a second data location in the substrate, the second seed location being separated from the second data location by the predetermined distance along the scan path, in the scan path the second seed location preceding the first seed location and the second data location, and wherein the second data location is identical to the first seed location, resulting in the formation of a birefringent voxel.

[0008] Related aspects provide a transparent optical data storage medium obtainable by the method.

[0009] Another aspect provides a system for forming a birefringent voxel in a transparent substrate. The system comprises: a pulsed laser source; a beam splitter arranged downstream of the pulsed laser source along an optical path; and focusing optics downstream of the beam splitter along the optical path. The beam splitter comprises: a retardation plate; a first polarization selective diffractive optic downstream of the quarter wave plate along the optical path; and a second polarization selective diffractive optic downstream of the first polarization selective diffractive optic along the optical path. The system can be used to implement the method described herein.

[0010] Yet another aspect provides a beam splitter. The beam splitter comprises: a retardation plate; a first polarization selective diffractive optic downstream of the retardation plate along an optical path; and a second polarization selective diffractive optic downstream of the first polarization selective diffractive optic along the optical path. The beam splitter can allow for generating split laser pulses with a controllable amplitude ratio.

[0011] This Summary is provided to introduce a selection of concepts in a simplified form that are further described below in the detailed description. This Summary is not intended to identify key features or essential features of the claimed subject matter, nor is it intended to be used to limit the scope of the claimed subject matter. The claimed subject matter is not limited to implementations that solve any or all disadvantages noted in any or all sections herein. BRIEF DESCRIPTION OF DRAWINGS

[0012] To aid understanding of embodiments of the present disclosure and to show how these embodiments function in practice, reference is made by way of example only to the accompanying drawings in which: Figure 1 is a schematic cross-sectional view of a birefringent optical data storage medium; Figure 2 is a flow chart outlining a method of forming birefringent voxels in a transparent substrate; Figures 3A to 3D shows an optical data storage medium after various iteration numbers of the method of Figure 2 Figure 4 is an illustration of a raster scan pattern used in the method of forming birefringent voxels; Figure 5 is a schematic plan view of an optical data storage medium obtainable following a scan path of Figure 4 Figure 6 is a block diagram of a write head for implementing the method of Figure 2 Figure 7 is an annotated schematic of a polarization grating based beam splitter; Figure 8A is a scanning electron microscope, SEM, image of a voxel obtained by a comparison process; Figure 8B is a SEM image of a voxel obtained by a method according to Figure 2 Figure 9 is a plot of data capacity versus number of laser pulses per voxel; Figure 10A is a plot of data capacity versus walk-off distance in a comparison method; Figure 10B is a plot of data capacity versus walk-off distance in a method according to Figure 2 Figure 11 is a plot of beam separation angle versus angle between two polarization gratings of a beam splitter according to Figure 7 Figure 12 is a plot of intensity ratio of two beams output by a beam splitter according to Figure 7 DETAILED DESCRIPTION

[0013] The structure of an example birefringent optical data storage medium will first be explained with reference to Figure 1 is a schematic cross-sectional view of an optical data storage medium. Figure 1

[0014] The data storage medium 100 comprises a substrate 110 having birefringent voxels 120 embedded therein.

[0015] ​​​​​​​​The substrate 110 can comprise glass, in particular fused silica. Fused silica can also be referred to as silica glass. Glass has excellent chemical and thermal stability and data storage media based on glass substrates have an expected lifetime of several hundred years. Thus, optical data storage media can be used for long-term archiving of data.

[0016] The shape and dimensions of the substrate are not particularly limited. Figure 1 A cuboid substrate is shown. In variants, the substrate can have any shape (e.g. a cylinder).

[0017] The thickness t of the substrate in the z-direction is typically at most 10 mm, optionally at most 5 mm, further optionally in the range of 200 pm to 2 mm. More generally, the thickness of the substrate can be appropriately chosen based on the capabilities of the imaging system to be used for imaging the voxels. The z-direction is the direction in which light will be transmitted through the substrate when imaging the voxels. The attenuation of light passing through the substrate increases with the distance travelled and providing a relatively thin substrate can limit such attenuation.

[0018] A plurality of voxels 120 is embedded in the substrate 110. A voxel is a discrete volume of the substrate which has been modified to have a non-natural birefringence, in other words a birefringence which is different from the birefringence of the bulk substrate material.

[0019] A voxel has a position which can be described by a set of coordinates x, y, z. A voxel can exhibit birefringence, in other words it can have different refractive indices for light of different polarizations. A voxel can cause a change in the angle and ellipticity of the polarization of light. A voxel has a linear retardation. The linear retardation is a measure of the magnitude of the phase shift between specific orthogonal linear polarization components of light after the light has passed through the voxel. The direction of the electric field oscillation of the specific orthogonal polarization component which gives the larger phase shift is referred to as the azimuth of the birefringence.

[0020] When a voxel is written into the substrate, the position, the retardation and the change in polarization angle of the voxel are controllable. Any of these properties or any combination of these properties can be used to encode digital data.

[0021] In the shown example, the voxels 120 are arranged as a stack of layers. The stack comprises a top layer 130, an intermediate layer 132 and a bottom layer 134. There can be any number of layers of voxels. For example, the optical data storage medium can comprise from 1 to 100 layers of voxels.

[0022] The voxels within a layer are laterally separated from each other. The distance between adjacent voxels within a layer can be referred to as the pitch p. Generally, the more closely the voxels are spaced, the greater the density of data that can be stored in the optical data storage medium per unit area.

[0023] To form a birefringent voxel in a substrate, at least two laser pulses are focused at a location in the substrate. Each laser pulse causes a photoexcitation of the material, causing a physical change of the substrate material at that location.

[0024] The substrate is most typically a transparent substrate. In this implementation, the voxel is typically embedded within the bulk of the substrate, but it is also possible to form the voxel at the surface of the substrate.

[0025] The use of an opaque or semi-transparent substrate is also contemplated. In such an implementation, the voxel is formed at the surface of the substrate.

[0026] A voxel formed on a surface can be referred to as a laser-induced periodic surface structure “LIPSS”.

[0027] The first pulse is referred to as a “seed pulse”. The seed pulse creates a seed modification, which can be a nanovoid or any other isotropic structure.

[0028] The second and any subsequent pulses are referred to as “data pulses”. The data pulse(s) are polarized laser pulse(s), focused on the location of the seed modification. The data pulse(s) modify the structure of the seed modification, thereby creating a birefringent voxel.

[0029] The data pulse(s) can have a linear, elliptical or circular polarization state. The birefringence of the voxel is determined by the polarization of the data pulse(s). For example, the azimuthal angle of the birefringence can be controlled by changing the polarization angle of the data pulse(s). This allows for polarization multiplexed optical data recording.

[0030] The seed pulse has a higher amplitude than the data pulses (in other words, is a higher energy laser pulse).

[0031] To avoid thermal damage to the substrate, a delay is required between the seed pulse and the data pulses.

[0032] There are two existing methods for forming birefringent voxels, referred to as “hot writing” and “cold writing”, respectively.

[0033] In the hot writing method, the seed pulse and the data pulses used to form a given voxel are applied sequentially through the same optical path. These steps are then repeated for each voxel to be written. To increase throughput, the substrate is moved during the writing process.

[0034] To successfully form a voxel, the seed pulse and the data pulses must be focused at points close enough to each other. The distance between the focal point of the seed pulse for a voxel and the focal point of the last data pulse for the voxel is referred to as the “walk-off distance”.

[0035] The acceptable "walk-off distance" constrains the speed at which the focal point of the write head can move, and requires precise relative positioning of the substrate to the write head. The "walk-off distance" in a thermal write method is given by the speed of movement of the focal point relative to the substrate divided by the pulse rate of the laser source. The acceptable walk-off distance is typically less than one-tenth of the voxel pitch.

[0036] A typical laser source operates at a uniform pulse rate in the tens of megahertz. The maximum speed of movement of the focal point relative to the substrate is slower compared to the laser pulse rate. Therefore, many laser pulses must be discarded. In a typical implementation of thermal writing, about 80% of the laser pulses are wasted.

[0037] Furthermore, because the seed pulse and the data pulse have different amplitudes, the thermal write process requires rapid modulation of the amplitude of the laser pulses. The rapid modulation generates significant signal noise and instability.

[0038] In a cold write method, the medium is prepared by forming a seed modification in the substrate in a first scan. Data pulses are then applied to the seed modification in a second scan.

[0039] The cold write method overcomes some of the limitations of thermal writing. Rapid modulation is not required, reducing hardware cost and eliminating one source of signal noise.

[0040] The cold write still has some limitations. Because two scans are required, the maximum possible throughput of the cold write method is half the pulse rate of the laser. Aligning the data pulses with the seed modification is challenging, and becomes increasingly difficult at faster scan rates. Furthermore, there is a substantial delay between forming the seed modification and applying the data pulses, making the technique susceptible to temporal variations in the calibration of the write system.

[0041] A method of forming birefringent voxels in a transparent substrate is provided herein that can address one or more limitations of existing methods.

[0042] The method is referred to as "pseudo-single-pulse writing," and involves splitting a single laser pulse into two or more spatially separated pulses with independently modulated amplitudes. The pulses are delivered to the transparent substrate simultaneously. At least one of the split pulses is used as a seed pulse, and at least another of the split pulses is used as a data pulse. These steps are iterated as a scan is made along a scan path in the transparent substrate. The data pulse of a later iteration reaches the location of a seed modification formed by an earlier iteration of the seed pulse, thereby forming a birefringent voxel.

[0043] The method can allow voxels to be written to the substrate at a rate approaching the pulse rate of the laser, thereby increasing throughput. The method can reduce the number of discarded laser pulses, thereby increasing hardware utilization efficiency and reducing energy consumption.

[0044] The method will now be explained with reference to the accompanying Figure 2 and Figures 3A to 3D The method is outlined in the flowchart of Figure 1. Figure 2 is a flowchart outlining the method. Figures 3A to 3D shows the state obtained after a series of iterations of the method.

[0045] The method is an iterative method. The number of iterations N s is a non-zero integer corresponding to the number of seed modifications to be made in the substrate. As will become apparent from the discussion below, N s is equal to the number of birefringent voxels to be written plus a non-zero integer constant.

[0046] Each iteration of the method comprises the operations of block 201 and block 202.

[0047] At block 201, a seed pulse S and a data pulse D are generated simultaneously. The operation of this block typically comprises generating a source laser pulse using a laser source, and splitting the source laser pulse into the seed pulse and the data pulse. The use of two or more laser sources generating synchronised pulses is also contemplated.

[0048] The laser source can be a femtosecond laser. The laser source can be a pulsed laser source generating laser pulses at a time interval t p The laser source generates laser pulses.

[0049] The properties of the beam splitter are not particularly limited and can be chosen as appropriate. The laser pulses can be split by passive beam splitting techniques or active beam splitting techniques. Example techniques include the use of a polarisation grating, a polarisation beam splitter, an acousto-optic deflector or a spatial light modulator.

[0050] The seed pulse and the data pulse have different respective amplitudes.

[0051] The seed pulse has an amplitude selected to form a seed modification in the transparent substrate. The seed modification can be any modification to the structure of the transparent substrate which results in a local change in the refractive index of the substrate compared to the refractive index of the original unmodified substrate. The seed modification is typically isotropic. In other words, the same refractive index, regardless of the polarisation of the light used to probe the seed modification.

[0052] Illustrative examples of seed modifications include forming densified or rarefied regions; local glass matrix changes; crystallisation or amorphisation; and forming nanoscale voids.

[0053] The seed pulse can be polarised or unpolarised.

[0054] The data pulse has an amplitude selected to modify the seed modification to induce birefringence. The amplitude of the data pulse is typically lower than the amplitude of the seed pulse.

[0055] The data pulses are polarized.

[0056] The method can further comprise modulating the polarization of the data pulses. For example, the data pulses can pass through a polarization state generator. By modulating the polarization of the data pulses, the birefringence properties of the voxel formed by the data pulses can be changed.

[0057] The polarization of the data pulses can be set individually for each iteration of the method.

[0058] In principle, more than one data pulse can be used, however using two or more data pulses per voxel does not provide a particular advantage. The inventors have found that the azimuth angle of the birefringent voxel is determined by the last data pulse to reach the voxel.

[0059] The seed pulse and the data pulse are separated by a predetermined distance in space upon reaching the transparent substrate.

[0060] The predetermined distance can have a fixed value for all iterations of steps 201 and 202. The predetermined distance can be jp V where j is an interval factor and p V is the inter-voxel distance in the scanning direction V f for the optical data storage medium. The interval factor j is a non-zero integer and can for example be in the range of 1 to 5.

[0061] To avoid or reduce cross-talk between the seed pulse and the data pulse, the predetermined distance is typically at least twice the wavelength of the laser light.

[0062] At block 202, the seed pulse and the data pulse are focused at respective locations in the transparent substrate. The seed pulse and the data pulse are separated by a predetermined distance jp v where the seed pulse is along the scanning direction V f preceding the data pulse.

[0063] The seed pulse and the data pulse can pass through the same optical components before reaching the transparent substrate. This can allow for a more consistent relative positioning of the seed pulse and the data pulse, as any variation in the operating parameters of the components will equally affect both the seed and data pulses.

[0064] Alternatively, the seed pulse and the data pulse can follow different optical paths. For example, the data pulse can pass through a polarization state generator and the seed pulse can bypass the polarization state generator. This can introduce a small difference (e.g. less than or equal to 1 m, more typically less than or equal to 10 cm) between the path lengths of the seed pulse and the data pulse. The pulses passing through different paths are still considered to arrive simultaneously, as the difference in arrival times is negligible compared to the speed of movement of the substrate relative to the write head.

[0065] Focusing the seed pulse at a location in the transparent substrate forms a seed modification at that location.

[0066] Focusing the data pulse at the location of the seed modification forms a birefringent voxel within a small tolerance, referred to as a walk-off distance. By way of illustration, a typical maximum walk-off distance can be as high as 100 nm.

[0067] Focusing the data pulse at a location other than the location of the seed modification can form a weak modification or no modification.

[0068] In the flowchart of Figure 2 at decision block 203, it is tested whether all desired birefringent voxels have been written into the transparent substrate. If so, the method proceeds to block 204 and terminates.

[0069] If not, the method performs further iterations of the steps of blocks 201 and 202. In further iterations, the focal points of the seed pulse and the data pulse are moved in the scan direction V f by a distance p v , which corresponds to the voxel spacing in the scan direction, as indicated by block 205.

[0070] Figures 3A to 3D An example state reached after a series of iterations of the above-described method is shown.

[0071] The value of the index i will be used to keep track of the iteration currently under discussion. The convention of zero indexing is followed. The value of i is incremented by one at each iteration. The use of an index can be convenient in implementations of the method in which the method is implemented as a computer-controlled process.

[0072] Figure 3A A first iteration (i = 0) of the method of Figure 2 is shown.

[0073] At time point t = 0, the write head 310, which can be a write head 600 as described below with reference to Figure 6 , focuses a data pulse D0 and a seed pulse S0 at respective locations in the substrate 320. As described previously, the data pulse and the seed pulse are generated simultaneously and delivered to the substrate.

[0074] The data pulse D0 reaches the unmodified portion of the substrate and forms a zeroth weak modification 330(0) in the substrate 320.

[0075] The seed pulse S0 forms a seed modification 340(0) at a location in the scan direction V f preceding the weak modification 330(0) by a distance jp v , recalling that p v will be the voxel spacing in the scan direction in the completed optical data storage medium, and j is a non-zero integer. In the example shown, the seed modification 340(0) is a birefringent voxel.Figures 3A to 3D In the example shown, j is 2.

[0076] Voxels also need to be written. This is done via box 205, where the writing head is aimed in the scanning direction V. f Upward shift distance p v Aiming the writing head may include adjusting the scanning optics of the writing head 310 and / or moving the substrate and / or the writing head relative to each other.

[0077] The method proceeds to the point where Figure 3B The next (first, i=1) iteration is shown.

[0078] The first iteration occurs at time t. p . t p This is the time interval between laser source-generated pulses written to the header 310. In other words, the first iteration uses a source laser pulse generated immediately after the source laser pulse used in the zeroth iteration.

[0079] according to Figure 2 In boxes 201 and 202, data pulse D1 and seed pulse S1 are generated and focused at corresponding positions in substrate 320.

[0080] Data pulse D1 targets the unmodified area of ​​the substrate and forms an additional weak modification 330(1). This additional weak modification 330(1) is in the scanning direction V. f The initial weak modification 330(0) downstream reaches the spacing p v .

[0081] Seed pulse S1 in scanning direction V f Up to the initial seed modified by 340(0), downstream spacing p v Another seed was created by modifying 340(1).

[0082] After this iteration, no voxels were formed. The write head was aimed in the scan direction V. f Increasing spacing p v And the method proceeds to Figure 3C The second (i=2) iteration is shown.

[0083] The second iteration (i=2) occurs at time 2t. p And use the laser pulse generated immediately after the laser pulse used in the first (i=1) iteration.

[0084] Next, the source laser pulse is split into a data pulse D2 and a seed pulse S2, which are aimed at a target position p farther away in the scanning direction than the previous D1 and S1 pulses. v The distance to the position.

[0085] The seed pulse S2 aims at a blank region of the substrate and forms a seed modification 340(2).

[0086] The index j for this iteration describes a factor that separates the data pulse from the seed pulse. The data pulse D2 reaches the location of the seed modification S0 from the zeroth (i = 0) iteration of the method. Instead of forming a weak modification 330, the data pulse D2 modifies the seed modification 340(0) to form a birefringent voxel 350(0). By modulating the polarization of the data pulse D2, the birefringence of the voxel 350(0) can be tuned as needed to encode a data symbol.

[0087] It will be appreciated that the data pulse used to form the voxel 350(0) reaches the seed modification at some time after the seed modification is formed. This allows for a time interval j p for the substrate to cool, thereby avoiding thermal damage to the substrate. For example, at a laser repetition rate of 50 MHz, the period between laser pulses t p is 20 ns. The separation factor j is 5, which provides a cooling time of 100 nanoseconds between the seed pulse and the data pulse reaching any given location.

[0088] The cooling time can be selected as appropriate based on the material selected and can be controlled by varying the separation factor j and / or the laser pulse separation t p . Of these parameters, the modification separation factor j can allow the cooling time to be adjusted without sacrificing data throughput.

[0089] The method continues with further iterations as long as voxels remain to be written.

[0090] A general iteration i of the method occurs at a time point it p from the laser source. The pulse from the laser source is split into a data pulse and a seed pulse. The data pulse is conveyed along the scan path V f to a location ip v unit. The seed pulse forms a seed modification at a location (i + j) p v unit along the scan path. If the iteration count i is greater than or equal to the separation factor j, a birefringent voxel is formed at a location (i - j) p f along the scan direction V v , otherwise the data pulse reaches a blank region and forms a weak modification.

[0091] The product of the final iteration is shown in Figure 3D .

[0092] The method forms an optical data storage medium 320 having j weak modifications in a data margin 360. The scan direction V fdownstream of the data margin on the substrate, there is a seed margin 380 comprising j non-birefringent seed modifications. s -2j) data array 370 of voxels.

[0093] Each voxel in the data array can be independently controlled by modulating the polarization of the corresponding data pulse to produce birefringence, thereby producing voxels encoding different data symbols as shown by the different shading patterns in Figure 3D

[0094] Downstream of the data array 370, there is a seed margin 380 comprising j non-birefringent seed modifications.

[0095] When the method is operating in steady state conditions, i.e. after the data margin has been formed and before the seed margin is started to be formed, the method can effectively use each pulse generated by the laser source to form a voxel. The overhead is small because j is small (e.g. in the range 1 to 5) compared to the number of voxels which do not have a specific upper limit.

[0096] Various modifications can be made to the illustrated example.

[0097] In this example, j is a non-zero integer. This can minimize the walk-off distance, thereby increasing the quality of the voxels and hence the number of bits that can be stored per voxel. In a variant, j can be a non-integer value, provided that the walk-off distance is not so large as to prevent the formation of a voxel.

[0098] The example has been explained with reference to writing a single row of voxels. Two or more rows can be written. When two or more rows are written, the rows can be written simultaneously in parallel.

[0099] The example shows relative motion of the write head and the substrate. More generally, any technique or combination of techniques can be used which allow the aim of the write head to be adjusted. The aim can be adjusted by: adjusting the orientation of a beam scanner of the write head; moving the write head; moving the substrate; and / or adjusting the focal depth of the write head.

[0100] The example shows writing of a single layer of voxels. Multiple layers of voxels can be written to increase the data storage capacity per unit volume.

[0101] In principle, the seed margin can be omitted by directing data pulses to additional seed modifications in the seed margin. However, this can involve modifications to the write head which can not be cost effective.

[0102] ​In this example, the scan path is linear. In variants, the scan can follow any suitable path. The scan path can be helical, particularly in implementations where the substrate is a cylinder or a disc. The scan path can be a raster scan path. The scan path can be a serpentine scan path, in other words, the scan path has a square wave shape. When following a scan path that includes reversing the scan direction, one or more optical components of the write head (e.g., a beamsplitter) are adjusted so that the seed pulse remains ahead of the data pulse.

[0103] This example shows the pulses separated in a direction perpendicular to the direction of laser propagation. In variants, the spatial separation can alternatively or additionally be in the direction of propagation. A polarizing lens can be used to achieve the splitting in the direction of propagation.

[0104] The described method can be parallelized, with several instances of the method occurring simultaneously. Parallelization can involve dividing the source laser pulse between two or more write paths using additional beamsplitters, with each write path configured to generate a seed pulse and a data pulse from the input pulse. Alternatively, each parallel write path can have a respective laser source.

[0105] Figure 4 An example raster scan is shown in Fig.

[0106] In the shown example, the substrate 420 is translated in the -x direction with a speed V s Continuously. The write head is moved along the y axis. The seed pulse and the data pulse propagate in the z direction.

[0107] Using the method explained with reference to Figure 2 and Fig. 3, the write head is scanned along a first linear scan path V f The write head is scanned to write a row of voxels. Then, the write head is moved in the +y direction to return to the original position, retracing the reset path R relative to the substrate 420. When the write head returns to the original position, no voxels are written. The scan pattern is repeated with a period T s This scan pattern is repeated until all desired voxels have been written.

[0108] This scan process forms a line of birefringent voxels, which can be referred to as a scan line. Adjacent scan lines are spaced apart from each other by a distance V s T s . The voxels within a scan line are spaced apart from each other by a pitch p V The pitch p v is the product of the period t p of the laser source and the scan speed for the row V f The spacing V s T s between scan lines can be different from the pitch p v between voxels within a scan line.

[0109] Due to the continuous movement of the substrate and the write head, the scan line is tilted V relative to the y-axis. s / V f Typically, V s / V f It's on the order of milliradians. This can be achieved by adjusting the substrate relative to the scanning direction V. s The angle can be used to reduce or eliminate the tilt.

[0110] pass Figure 4 The optical data storage medium 520 obtained during the scanning process is in Figure 5 As shown in the figure. Ideally, voxels would be arranged in uniformly spaced scan lines, with each scan line at a consistent angle relative to the substrate.

[0111] As can be seen, the geometry of the scan lines of voxel 550 may deviate from the ideal. Factors such as tolerances in mechanical alignment, drift in the characteristics of optical components (e.g., due to temperature variations), and fluctuations in the laser source can cause voxel displacement.

[0112] The effects of any such displacement can be mitigated by forming a predetermined reference mark in the substrate. The reference mark can also be referred to as a preamble.

[0113] A reference marker is a set of voxels arranged in a predetermined pattern. Typically, the pattern is a one-dimensional (“1D”) or two-dimensional (“2D”) pattern, but a 3D pattern may be used alternatively. There is no particular limitation on the size of the set, as long as the number of voxels is large enough to allow the pattern to be clearly identified when reading optical data storage media. For illustration, a 2D reference marker can be at least 4 voxel widths multiplied by 4 voxel heights.

[0114] When processing images from optical data storage media to recover stored data, reference markers can be identified, and the positions of the voxels that form these reference markers can be determined. Based on the determined positions, corrections can be applied to compensate for positional changes that occur when any distortion is written to the voxels and / or the image.

[0115] Location information can be used as input to processing / decoding methods for recovering data from optical data storage media. These processing / decoding methods may include the use of machine learning models.

[0116] Reference markers may be arranged, for example, along one or more edges of a sector of a voxel. Other arrangements are conceivable. For example, reference markers may alternatively or additionally be arranged at the center of a sector.

[0117] A sector is a two-dimensional group of voxels. In many implementations, a given layer of voxels in an optical data storage medium comprises two or more sectors, and is read using a reader whose field of view is sized to capture an image comprising one complete sector and up to 8 edge portions of immediately adjacent sectors.

[0118] The predetermined pattern can be a Barker sequence or a Frank-Zadoff-Chu sequence. Other patterns can be used.

[0119] The symbols forming the fiducial marks can be selected so as to maximize the contrast between the symbols. This can allow the marks to be more easily detected.

[0120] Two or more different fiducial marks can be present in a single optical data storage medium. Adjacent sectors can be associated with different fiducial marks. This can allow sectors to be more easily distinguished from one another.

[0121] For example, two different fiducial marks can be used, with the associated sectors arranged in a checkerboard pattern so that no sector shares an edge with another sector having the same fiducial mark.

[0122] When writing an optical data storage medium using a high-throughput method, such as the method described with reference to Figure 2 It is particularly useful to write fiducial marks into an optical data storage medium when writing the optical data storage medium using a high-throughput method, such as the method described with reference to

[0123] An example laser writing system 600 will now be described with reference to Figure 6 The laser writing system 600 is used to implement the methods described herein. Figure 6 is a schematic block diagram of the laser writing system 600.

[0124] The example laser writing system 600 comprises a laser source 610, a first beamsplitter 630, a write path 640 and a sample stage 660. The diagram shows the laser writing system 600 during use, with a substrate 620 mounted on the sample stage.

[0125] The laser source 610 generates laser pulses. Typically, the laser pulses are generated with a constant repetition rate having a period t p As an illustration, the repetition rate can be of the order of 10 MHz. The laser source 610 can be a femtosecond laser.

[0126] The laser source 610 emits laser pulses toward the first beamsplitter 630. The first beamsplitter 630 splits each pulse into two or more beams. Each beam is directed toward a respective write path 640. Splitting the laser into two or more beams allows two or more instances of the methods described herein to be performed in parallel, for example to allow two or more scan lines to be written simultaneously.

[0127] A single write path 640 is shown in the figure. The write path 640 comprises in sequence an intensity modulator 642, a second beamsplitter 644, a polarization modulator 646, a scanner 648 and an objective lens 650.

[0128] The intensity modulator 642 receives laser pulses from the first beamsplitter 630 and adjusts the intensity of the laser pulses. The intensity of a voxel can vary according to the amplitude of the laser pulse used to form the voxel, and the intensity modulator can allow this to be adjusted.

[0129] The amplitude modulator can comprise an intensity attenuator and / or an amplifier.

[0130] The intensity attenuator may, for example, comprise an acousto-optic deflector; a liquid crystal modulator; a combination of electro-optic modulator and polarizer; or a combination of Pockels cell and polarizer. These are examples of fast amplitude modulators. Amplitude modulators capable of pulse-to-pulse energy control are considered to be "fast".

[0131] The intensity modulator 642 can be an acousto-optic deflector or a combination of electro-optic modulator and polarizer.

[0132] The laser pulses then reach the second beamsplitter 644. As previously described, the second beamsplitter 644 splits the pulses into seed pulses and data pulses having different respective intensities. The second beamsplitter can be an electro-optic deflector, an acousto-optic deflector or a polarization grating-based beamsplitter.

[0133] The second beamsplitter 644 can alternatively comprise a passive beamsplitter (for example, selected from a non-polarization sensitive beamsplitter, a polarization sensitive beamsplitter, a diffractive optical element, a spatial light modulator) in combination with an active spatial modulator and an active amplitude modulator.

[0134] The second beamsplitter 644 is most preferably a beamsplitter of the type described below with reference to Figure 7

[0135] At least the data pulses from the second beamsplitter 644 pass through the polarization modulator 646. The polarization modulator modulates the polarization of the data pulses at the laser repetition rate.

[0136] ​The polarization modulator 646 can be, for example, a polarization state generator. Example polarization state generators include a linear polarizer and a quarter waveplate arranged in series. Other examples of polarization state generators include a Pockels cell; a magneto-optic modulator; a liquid crystal; a rotating prism; a rotating periscope; etc.

[0137] Since no particular polarization of the seed pulse is required, the seed pulse can pass through the polarization modulator 646, or alternatively can bypass the polarization modulator 646. In implementations in which the seed pulse passes through the polarization modulator 636, the polarization modulator 646 modulates the polarization of both the data pulses and the seed pulse.

[0138] A scanner 648 is arranged downstream of the polarization modulator 646. The scanner 648 can include a rotating polygon mirror, a micro-electro-mechanical system ("MEMS") mirror, a Galvano scanner, an electro-optic scanner, an acousto-optic scanner, etc.

[0139] In implementations in which the seed pulse does not pass through the polarization modulator 646, the scanner 648 can be aligned to the seed pulse and the data pulses.

[0140] The scanner 648 deflects the seed pulse and the data pulses. The deflection angle can be swept at a constant velocity.

[0141] The scanner 648 outputs the seed pulse and the data pulses to an objective lens 650. The objective lens 650 focuses the seed pulse and the data pulses on or in the substrate 320 to allow for the formation of voxels, as described with reference to Figure 2 .

[0142] The objective lens 650 can have a variable focal depth or focal position. Voxels can then be written at variable depth positions within the transparent substrate 620 without requiring the transparent substrate to move in the z-direction.

[0143] A relay optic can be arranged between the scanner 648 and the objective lens 660. The relay optic can include a scan lens, a spherical lens, and a tube lens.

[0144] The laser writing system also includes a sample stage 650 for holding the substrate 620. The sample stage can be a translation stage that moves the substrate 620 perpendicular to the beam direction and perpendicular to the beam scan direction.

[0145] Various modifications can be made to the example write head.

[0146] The first beam splitter 620 is optional and can be omitted in implementations that do not use parallelization of the method.

[0147] There is no particular upper limit to the number of write paths in case parallelization is used. In case more than two write paths are to be used, the first beamsplitter 620 can be replaced by two or more beamsplitters arranged in series. As an illustration, the system can comprise 1 to 5 write paths.

[0148] In an example, the scanner 648 and the objective 650 are components of the write path. Alternatively, the scanner 648 and the objective 650 can be shared between several write paths. In such an implementation, the light beams can be recombined after the polarization is modulated.

[0149] The intensity modulator 642 is optional and can be omitted.

[0150] In principle, the write path can comprise separate optical components for the seed pulse and the data pulse, respectively. It can be advantageous for the seed pulse and the data pulse to follow the same optical path to allow for more accurate control of the distance between the seed pulse and the data pulse.

[0151] The write head can comprise one or more additional optical relay components (e.g., one or more lenses) arranged between the second beamsplitter 644 and the scanner 648, such that the seed pulse and the data pulse reach the same region of the scanner.

[0152] The scanner 648 can be omitted. The positions of the data pulse and the seed pulse in the transparent substrate can instead be controlled only by movement of the sample stage.

[0153] The example sample stage is a translation stage. Alternatively or additionally, the sample stage can rotate the transparent substrate.

[0154] The illustrated example uses a single laser source to generate both the seed pulse and the data pulse. In other implementations, the seed pulse and the data pulse can be generated by respective laser sources. In such implementations, the second beamsplitter can be omitted. When two or more laser sources are used, the laser sources can be laser sources of different wavelengths.

[0155] Reference will now be made to Figure 7 An example beamsplitter 700 is described. Figure 7 is an annotated schematic of the beamsplitter 700.

[0156] The beamsplitter 700 comprises a delay plate 710, a first polarization-selective diffractive optic 720, and a second polarization-selective diffractive optic 730 arranged in series.

[0157] The polarization-selective diffractive optics can alternatively be referred to as “geometrical phase optical elements” or “geometrical phase elements”.

[0158] In this example, the retardation plate 710 is a quarter waveplate; the first polarization selective diffractive optic 720 is a first polarization grating; and the second polarization selective diffractive optic is a second polarization grating.

[0159] The retardation plate 710, the first polarization grating 720, and the second polarization grating 730 are arranged in series.

[0160] In use, polarized input light LP (e.g., originating from the laser source 610 of the laser writing system 600) reaches the quarter waveplate 710. The input light can be linearly polarized, elliptically polarized, or circularly polarized. For ease of explanation, in the following discussion, the input light will be considered to be linearly polarized light.

[0161] When the linearly polarized light passes through the retardation plate 710, it becomes elliptically polarized light EP. Any degree of ellipticity can be obtained by adjusting the angle A of the quarter waveplate 710.

[0162] The elliptically polarized light EP then reaches the first polarization grating 720. The first polarization grating 720 splits the elliptically polarized light into a left circularly polarized beam and a right circularly polarized beam. The relative amplitudes of the two beams are determined by the ellipticity of the elliptically polarized light EP. Thus, the polarization grating-based beam splitter 700 is a convenient component for generating seed and data pulses of an intended amplitude ratio in a method of Figure 2

[0163] The left and right circularly polarized beams then reach the second polarization grating 730. Since the beams are circularly polarized, each beam is deflected in a single direction by the second polarization grating.

[0164] The angle Θ between the beams LHC, RHC OUT The angle 1 of the first polarization grating can be changed by changing the angle 2 of the second polarization grating with respect to it.

[0165] In more detail, the deflection angle can be changed by changing the relative orientation of the grating vectors of the first and second polarization gratings. When the grating vectors are parallel, the beam splitting angle is minimized; while when the grating vectors are anti-parallel, the beam splitting angle is maximized.

[0166] The mathematical treatment of the beam splitting will now be provided.

[0167] Elliptical polarization can be represented by a Jones vector:

[0168] where c is the ellipticity.

[0169] Right and left hand circular polarization is represented as:

[0170] ​where + and - are left and right handed, respectively.

[0171] Equation A1.1 can be expressed as a linear combination of the Jones vectors of the two circular polarizations:

[0172] Using beamsplitter 700, the splitting angle can be tuned by changing the relative orientation of polarization gratings 720, 730.

[0173] For illustration, first consider the case where the input light EP is right circularly polarized light. This light is deflected in one direction only by the first polarization grating 720, and is converted to left circularly polarized light. The spatial phase change induced by the first polarization grating is:

[0174] where q1is the deflection angle of the first polarization grating, j1is the tilt of the grating vector of PG1with respect to the x-axis, and l is the wavelength of the beam.

[0175] By symmetry, the second polarization grating 730 induces the following phase change:

[0176] where q2is the deflection angle of the second polarization grating, and j2is the tilt of the grating vector of PG2with respect to the x-axis.

[0177] Thus, the total phase change for right circularly polarized light passing through the beamsplitter is:

[0178] When the deflection angles of the two PGs are the same (q = q1= q2), then equation A2.3 can be simplified to:

[0179] The overall phase change for left circularly polarized light passing through the beamsplitter is:

[0180] Elliptically polarized light can be viewed as a combination of left and right circularly polarized light. According to equation A1.3, an ellipse polarization with an ellipticity of c has two circular polarizations with an energy ratio of:

[0181] Thus, the left and right circularly polarized components experience different phase changes (equations A2.4 and A2.5). These phase changes result in the light being split at the following angles:

[0182] The tilt of the beamsplitting plane with respect to the xy-plane is:

[0183] Equations A2.7 and A2.8 show that by changing the relative orientation (j1 - j2) between the two polarization gratings from 0 degrees (parallel) to 180 degrees (anti-parallel), the splitting angle can be changed from 0 to about 4q, where the splitting plane is determined by the sum of the orientations (j1 + j2). Furthermore, Equation A2.6 indicates that the amplitude ratio of the two beams can be changed by changing the ellipticity of the input light.

[0184] Various modifications can be made to the described polarization grating based beam splitter.

[0185] For example, the quarter wave plate 710 can be replaced by a wave plate with a different retardation, e.g. a third wave plate. A third wave plate can provide a smaller ratio between the amplitudes of the light beams output by the beam splitter.

[0186] As an alternative to polarization gratings, polarization holograms or polarization lenses can be used. For example, when the polarization gratings are replaced by polarization lenses, the beam splitter can generate two differently diverging light beams. The differently diverging light beams can be combined, e.g. using the focusing optics 650 of the laser writing system 600, to generate a multi-focal light beam (in other words, two focussed light beams in the propagation direction of the light).

[0187] Two or more beam splitters of the described type can be arranged in series to generate multiple light beams with variable propagation directions and variable amplitude ratios.

[0188] Example 1 A birefringent voxel layer is written into a silica glass medium using the method of Figure 2 . The laser source of the laser writing system is a femtosecond laser. The polarization modulator is synchronized with the laser. The laser writing system further comprises a high speed beam scanner, an XYZ translation stage, a fast attenuator and a beam splitter.

[0189] The writing is performed using a laser writing system of the type described with reference to Figure 6 . The laser source of the laser writing system is a femtosecond laser. The polarization modulator is synchronized with the laser. The laser writing system further comprises a high speed beam scanner, an XYZ translation stage, a fast attenuator and a beam splitter.

[0190] The data is encoded into 8 polarization azimuth angle values, i.e. 3 bits per voxel. One seed pulse and one data pulse are used to write each voxel with an intensity ratio of about 0.6 to 0.7.

[0191] The lateral pitch is about 0.5 x 0.5 micrometer. The spatial separation between the seed beam and the data beam inside the medium is about 2 micrometers. The beam scanning speed is about 5 m / s inside the medium.

[0192] A peak throughput of about 10 million voxels per beam per second is achieved.

[0193] Example 2 Two types of optical data storage voxels were written using the comparative thermal writing method and the method according to Figure 2 Lei et al., Optica 8, no. 11, pp. 1365-1371 (2021), respectively. Scanning electron microscope images of voxels in the medium are shown in Figure 8A and Figure 8B , respectively.

[0194] It was found that the voxels obtained by the comparative method comprise a central protrusion and asymmetric elongated cracks on either side of the central protrusion Figure 8A . In contrast, the voxels obtained by the method according to Figure 2 have an elliptical shape Figure 8B .

[0195] Without wishing to be bound by theory, it is believed that the difference in shape can be caused by the improved alignment between the seed pulse and the data pulses achieved by the method according to Figure 2 . The asymmetry of the voxels obtained using the comparative thermal writing method is believed to be caused by the seed pulse and the data pulses arriving at slightly different locations, which is unavoidable in the thermal writing process.

[0196] Example 3 Data was written to an optical storage medium using the cold writing method. A different number of data pulses per voxel was used. The storage capacity of the medium in bits per voxel was measured. The results are shown in Figure 9 .

[0197] It was found that a difference in polarization between the seed pulse and the data pulses has no significant effect on the quality of the voxels Figure 9 , the two leftmost data points. The azimuthal angle of birefringent voxels is determined by the polarization of the data pulse(s) and is not sensitive to the polarization of the seed pulse.

[0198] In the case where more than one data pulse per voxel was used (i.e. the total number of pulses per voxel was greater than 2), a marginal improvement in quality was observed when all data pulses had the same polarization. The improvement was not large enough. When the data pulses had different polarizations, the quality decreased.

[0199] Thus, Figure 9 it was shown that the quality of birefringent voxels is not affected by the seed polarization and that there is no meaningful benefit to using more than one data pulse per voxel.

[0200] Example 4 The effect of walk-off distance on the data capacity per voxel was investigated for the comparative cold writing process and for the process according to Figure 2 .

[0201] Walk-off of more than ±20 nm was found to cause observable degradation of birefringent voxels in a cold write process Figure 10A ).

[0202] When using the process according to Figure 2 ( Figure 10B ), a similar effect was observed when the seed pulse and the data pulse were separated by a non-integer multiple of the voxel pitch.

[0203] The results show that the separation between the seed pulse and the data pulse is most preferably set to an integer multiple of the voxel pitch to minimize the walk-off distance and thereby maximize the data capacity.

[0204] Example 5 A reference Figure 7 splitter of the type described was constructed. In the example device, the two polarization gratings had the same grating period.

[0205] The orientation of the polarization gratings (“PGs”) was varied. More specifically, the difference between the orientations was varied while keeping their sum the same . The distance between the two output beams was monitored using a beam profiler. The separation angle was calculated based on the measured distance. The results are shown in Figure 11 .

[0206] The maximum separation angle in this measurement was 3.7 mrad, which is twice the separation angle achieved using a single PG. The separation angle follows the equation:

[0207] where 2q0is the separation angle for two different circular polarizations by a single PG.

[0208] Additional experiments were performed to demonstrate that the relative amplitudes of the two beams output by the splitter can be changed by adjusting the angle of the quarter wave plate (“QWP”). Linearly polarized light was used as input. The results are shown in Figure 12 .

[0209] It was found that when the polarization before the QWP is linear, the intensity ratio follows the sine of the angle of the QWP:

[0210] where q QHP is the orientation of the QWP, and the angle of the input polarization is 45 degrees, i.e., the polarization after the QWP is in q QHP = 0°, 180° RHCP.

[0211] It will be appreciated that the above described embodiments are described by way of example only.

[0212] More generally, according to one aspect disclosed herein, there is provided a method of forming a birefringent voxel in a transparent substrate. The method comprises the steps of: i) simultaneously generating a first seed pulse and a first data pulse, the first seed pulse and the first data pulse being spatially separated laser pulses, the first seed pulse having an amplitude different from an amplitude of the first data pulse; ii) focusing the first seed pulse at a first seed location in the substrate and focusing the data pulse at a first data location in the substrate, the first seed location being separated from the first data location by a predetermined distance along a scan path, wherein in the scan path the first seed location precedes the first data location; and iii) after steps i) and ii), simultaneously generating a second seed pulse and a second data pulse, the second seed pulse and the second data pulse being spatially separated laser pulses, the second seed pulse having an amplitude different from an amplitude of the second data pulse; iv) focusing the second seed pulse at a second seed location in the substrate and focusing the data pulse at a second data location in the substrate, the second seed location being separated from the second data location by a predetermined distance along a scan path, in which the second seed location precedes the first seed location and the second data location, and wherein the second data location is the same as the first seed location, resulting in the formation of a birefringent voxel.

[0213] A related method comprising the steps of: i) simultaneously generating a seed pulse and a data pulse, the seed pulse and the data pulse being spatially separated laser pulses, and the amplitude of the seed pulse being different from the amplitude of the data pulse; ii) focusing the seed pulse and the data pulse at respective locations in a substrate, the locations being separated by a predetermined distance along a scan path, wherein in the scan path the location of the seed pulse precedes the location of the data pulse; and iii) iterating steps i) and ii), wherein the locations of the seed pulse and the data pulse are incremented along the scan path between iterations, the location of the seed pulse in an earlier iteration being the same as the location of the data pulse in a later iteration, resulting in the formation of a birefringent voxel at that location. It will be appreciated that the various optional features described below can be incorporated into implementations of either method.

[0214] Typically, the amplitude of the seed pulse is greater than the amplitude of the data pulse.

[0215] The substrate can be a transparent substrate. “Transparent” means optically transparent to the wavelengths of the seed pulses and the data pulses. When the substrate is a transparent substrate, the locations of the seed pulses and the data pulses can be within the bulk of the substrate or at the surface of the substrate.

[0216] Alternatively, the substrate can be an opaque substrate. In such an implementation, the locations are at the surface of the substrate.

[0217] Generating the seed pulses and the data pulses can comprise generating a source laser pulse and splitting the source laser pulse using a beamsplitter. This can allow for a more efficient use of the laser.

[0218] Alternatively, the seed pulses and the data pulses can be generated by respective laser sources. This can allow the seed pulses and the data pulses to have different wavelengths.

[0219] The beamsplitter can comprise a retardation plate, a first polarization-selective diffractive optic downstream of the retardation plate along an optical path, and a second polarization-selective diffractive optic downstream of the first polarization-selective diffractive optic along the optical path. Such a beamsplitter can allow for controlling the amplitudes of the source pulses and the data pulses as well as a predetermined distance.

[0220] The retardation plate can be a quarter-wave plate.

[0221] Examples of polarization-selective diffractive optics include polarization gratings, polarization holograms, and polarization lenses. The first polarization-selective diffractive optic and the second polarization-selective diffractive optic can be the same or different. In particular, the first polarizer can be a first polarization grating and the second polarizer can be a second polarization grating.

[0222] The locations of the seed pulses and the data pulses can be incremented by a non-integer multiple of a predetermined distance between iterations. In other words, the locations of the seed pulses and the data pulses can be incremented by a non-integer multiple of a voxel spacing. This can allow the data pulses of later iterations to be more reliably aligned with the seed pulses of earlier iterations.

[0223] The non-integer multiple can be, for example, at least 2. Changing the spacing between the seed pulses and the data pulses changes the time delay between the arrival of a seed pulse and the arrival of a data pulse at any given location. Providing such a time delay can avoid thermal damage to the transparent substrate.

[0224] The predetermined distance can be greater than or equal to twice the wavelength of the seed pulses and the data pulses. This can reduce or avoid cross-talk between the seed pulses and the data pulses. Smaller predetermined distances can be used.

[0225] The method can further comprise modulating the polarization of the data pulses. In other words, the polarization of the data pulses of different iterations can be different. Modulating the polarization can allow for high density storage of data, as each different polarization can represent a different data symbol.

[0226] The seed pulses and the data pulses can share a common path through the scanning optics and the transfer optics. This can allow for more reliable positioning of the seed pulses and the data pulses.

[0227] The iteration steps i) and ii) can result in the formation of a plurality of birefringent voxels, wherein the plurality of birefringent voxels represents a fiducial marker. The fiducial marker is most typically a 1 -dimensional or 2-dimensional fiducial marker, but 3-dimensional fiducial markers can also be used. When recovering data from the optical data storage medium formed by the method, a reader can identify the fiducial marker. This can allow the reader to compensate for positional inaccuracies during the writing process.

[0228] The method can be parallelized. Two or more instances of the method can be executed simultaneously. Two or more instances of steps i) and ii) can be executed simultaneously in parallel. This can increase throughput.

[0229] Another aspect provides an optical data storage medium that can be obtained by the method. As explained herein with reference to Figure 8A and Figure 8B The voxels written by the present method have a different structure than the voxels produced by a comparative thermal writing process.

[0230] The optical data storage medium can be a transparent optical data storage medium comprising a transparent substrate and birefringent voxels embedded in the transparent substrate. The birefringent voxels can be in the form of an elliptical void having a diameter as measured along the semi-major axis of the ellipse in the range of 1 to 200 nm, optionally in the range of 50 to 200 nm.

[0231] The optical data storage medium can comprise voxels arranged in a scan line. A first end region of the scan line can comprise at least one weak modification, which can be obtained by applying a data pulse to a blank region of the substrate. A second end region of the scan line can comprise one or more non-birefringent seed modifications. A data encoding region comprising one or more birefringent voxels can be arranged between the first and second end regions.

[0232] The optical data storage medium can comprise layers of two or more such scan lines. The spacing between the voxels within a scan line can be different from the spacing between adjacent scan lines.

[0233] The optical data storage medium can comprise two or more such layers arranged in a stack. The number of layers is not particularly limited, provided that data is recoverable by a suitable reading system. By way of illustration, the optical data storage medium can comprise 2 to 100 layers or more of birefringent voxels.

[0234] The optical data storage medium can comprise a plurality of birefringent voxels arranged as fiducial marks (optionally, two-dimensional fiducial marks). The two-dimensional fiducial marks can be used to assist in recovering data encoded by the voxels. For example, the positions of the voxels in the two-dimensional fiducial marks can be used to determine one or more correction parameters for compensating for errors in the positioning of the voxels.

[0235] The birefringent voxels can be arranged in a plurality of sectors. Each sector can be associated with a respective fiducial mark, for example arranged at an edge of the sector.

[0236] The transparent substrate can be a glass substrate, optionally a silica glass substrate or a quartz glass substrate. Glass has excellent chemical, thermal and physical stability, allowing very long-term storage of data.

[0237] Another aspect provides a laser writing system for implementing the method. The laser writing system comprises: a laser source for emitting laser pulses; and a write path for receiving input laser pulses originating from the laser source and focusing seed pulses and data pulses at respective locations in a substrate.

[0238] The write path comprises: a beam splitter for splitting the laser pulses into seed pulses and data pulses, the seed pulses and data pulses being spatially separated, the amplitude of the seed pulses being different from (e.g. greater than) the amplitude of the data pulses; and an objective lens for focusing the seed pulses and data pulses at different respective locations in the substrate, the objective lens being downstream of the beam splitter in the optical path.

[0239] The write path can further comprise scanning optics, for example arranged between the beam splitter and the objective lens. The scanning optics can change the positions of the seed pulses and data pulses.

[0240] The laser writing system can further comprise or be operatively linked to a controller configured to control the write head to perform the method as described herein. The nature of the controller is not particularly limited.

[0241] The controller can comprise a dedicated hardware circuit, for example an application specific integrated circuit ASIC. The controller can comprise one or more processors and a non-transitory computer readable data storage device having stored therein instructions which, when executed by the processor(s), cause the write head to perform the method as described herein. A combination of dedicated hardware circuitry and one or more processors can be used.

[0242] The beam splitter can comprise: a retardation plate; a first polarization-selective diffractive optic downstream of the retardation plate along the optical path; and a second polarization-selective diffractive optic downstream of the first polarization-selective diffractive optic along the optical path. Such a beam splitter provides a convenient means for generating seed pulses and data pulses having respective different amplitudes and a predetermined separation.

[0243] The retardation plate can be a quarter-wave plate.

[0244] Examples of polarization-selective diffractive optics include polarization gratings, polarization holograms, and polarization lenses. The first and second polarization-selective diffractive optics can be the same or different. In particular, the first polarization-selective diffractive optic can be a first polarization grating, and the second polarization-selective diffractive optic can be a second polarization grating.

[0245] The beam splitter can be configured to allow relative rotation of the first and second polarization-selective diffractive optics. This can allow control of the separation between the seed pulses and the data pulses.

[0246] The retardation plate can be adjustable. Adjusting the retardation plate can change the ellipticity of the light reaching the first polarization-selective diffractive optic, and in turn change the amplitude ratio of the seed pulses and the data pulses.

[0247] The laser writing system can further comprise a polarization modulator for modulating the polarization of the data pulses. Modulating the polarization of the data pulses allows voxels having different birefringence to be written into the transparent substrate. Differences in birefringence (e.g. azimuthal angle) can be used to encode data.

[0248] The laser writing system can comprise two or more writing paths. Providing two or more writing paths can allow two or more instances of the methods provided herein to be performed simultaneously.

[0249] For example, the laser writing system can comprise a further beam splitter arranged downstream of the laser source and upstream of the writing paths. The second beam splitter can provide the input pulses to each of the two or more writing paths.

[0250] Another example laser writing system comprises: a pulsed laser source; a beam splitter arranged downstream of the pulsed laser source along an optical path; and a focusing optic downstream of the beam splitter along the optical path. It will be appreciated that this laser writing system can comprise any of the features of the previously described systems. For example, the beam splitter can comprise: a retardation plate; a first polarization-selective diffractive optic downstream of the quarter-wave plate along the optical path; and a second polarization-selective diffractive optic downstream of the first polarization-selective diffractive optic along the optical path.

[0251] In another aspect, there is provided a beamsplitter. The beamsplitter comprises: a retardation plate; a first polarization-selective diffractive optic downstream of the retardation plate along an optical path; and a second polarization-selective diffractive optic downstream of the first polarization-selective diffractive optic along the optical path. The beamsplitter can be used to split a beam of polarized light. As described above, such a beamsplitter allows the relative amplitudes and angular separation of the split beams to be controlled.

[0252] The retardation plate can be a quarter-wave plate.

[0253] Examples of polarization-selective diffractive optics include polarization gratings, polarization holograms, and polarization lenses. The first and second polarization-selective diffractive optics can be the same or different. In particular, the first polarization-selective diffractive optic can be a first polarization grating, and the second polarization-selective diffractive optic can be a second polarization grating.

[0254] The first and second polarization-selective diffractive optics can be rotatable relative to each other. This can allow the angular separation between the split beams to be varied.

[0255] The retardation plate can be adjustable, e.g. rotatable. Adjusting the retardation plate can adjust the ellipticity of the polarized light reaching the first polarization-selective diffractive optic, in turn varying the relative amplitudes of the split beams produced by the beamsplitter.

[0256] The present disclosure provides the following clauses: Clause 1. A method comprising: generating simultaneously a first seed pulse and a first data pulse, the first seed pulse and the first data pulse being spatially separated laser pulses, and the amplitude of the first seed pulse being different from the amplitude of the first data pulse; focusing the first seed pulse at a first seed location in a substrate and focusing the data pulse at a first data location in the substrate, the first seed location being separated from the first data location by a predetermined distance along a scan path, wherein in the scan path, the first seed location precedes the first data location; after focusing the first seed pulse at the first seed location and focusing the data pulse at the first data location, generating simultaneously a second seed pulse and a second data pulse, the second seed pulse and the second data pulse being spatially separated laser pulses, and the amplitude of the second seed pulse being different from the amplitude of the second data pulse; and focusing a second seed pulse at a second seed location in the substrate and focusing a data pulse at a second data location in the substrate, the second seed location being separated from the second data location by a predetermined distance along a scan path in which the second seed location precedes the second data location, and in which the second data location is the same as the first seed location, thereby causing formation of a birefringent voxel.

[0257] Clause 2. The method of clause 1, wherein the first seed pulse has an amplitude that is greater than an amplitude of the first data pulse.

[0258] Clause 3. The method of clause 1 or clause 2, wherein the substrate is a transparent substrate.

[0259] Clause 4. The method of any of the preceding clauses, wherein generating the first seed pulse and the first data pulse comprises: generating a source laser pulse, and splitting the source laser pulse using a beam splitter.

[0260] Clause 5. The method of clause 4, wherein the beam splitter comprises: a retardation plate; a first polarization-selective diffractive optic downstream of the retardation plate along an optical path; and a second polarization-selective diffractive optic downstream of the first polarization-selective diffractive optic along the optical path.

[0261] Clause 6. The method of clause 5, wherein the retardation plate is a quarter- wave plate.

[0262] Clause 7. The method of clause 5 or clause 6, wherein the first polarization- selective diffractive optic and the second polarization-selective diffractive optic are each independently selected from the group consisting of a polarization grating, a polarization hologram, and a polarization lens.

[0263] Clause 8. The method of any of the preceding clauses, further comprising, after focusing the first seed pulse at the first seed location and focusing the data pulse at the first data location, and before focusing the second seed pulse at the second seed location and focusing the data pulse at the second data location, performing at least one iteration of: a) simultaneously generating an intermediate seed pulse and an intermediate data pulse, the intermediate seed pulse and the intermediate data pulse being spatially separated laser pulses of different amplitudes; b) focusing the intermediate seed pulse and the intermediate data pulse at an intermediate seed and an intermediate data location in the substrate, respectively, the intermediate seed location and the intermediate data location being separated by the predetermined distance, and in the scan path, the intermediate seed location preceding the intermediate data location.

[0264] Clause 9. The method of clause 8, wherein adjacent seed positions are separated by a non-integer divisor of the predetermined distance between iterations.

[0265] Clause 10. The method of clause 9, wherein the non-integer divisor is at least 2.

[0266] Clause 11. The method of any of the preceding clauses, wherein the predetermined distance is greater than or equal to twice the wavelength of the seed pulses and the data pulses.

[0267] Clause 12. The method of any of the preceding clauses, further comprising modulating the polarization of the first data pulse and the second data pulse.

[0268] Clause 13. The method of any of the preceding clauses, wherein the first seed pulse and the first data pulse share a common path through a scanning optic and a delivery optic.

[0269] Clause 14. The method of any of the preceding clauses, further comprising: generating a third seed pulse and a third data pulse simultaneously with the first seed pulse and the first data pulse, the third seed pulse and the third data pulse being spatially separated laser pulses and the amplitude of the third seed pulse being different from the amplitude of the third data pulse; focusing the third seed pulse at a third seed position in the substrate and the third data pulse at a third data position in the substrate simultaneously with focusing the first seed pulse at the first seed position and the first data pulse at the first data position, the third seed position and the third data position being separated by the predetermined distance along another scan path, wherein on the other scan path, the third seed position precedes the third data position.

[0270] Clause 15. A method comprising the steps of: i) generating a seed pulse and a data pulse simultaneously, the seed pulse and the data pulse being spatially separated laser pulses and the amplitude of the seed pulse being different from the amplitude of the data pulse; ii) focusing the seed pulse and the data pulse at respective positions in a substrate, the positions being separated by a predetermined distance along a scan path, wherein on the scan path, the position of the seed pulse precedes the position of the data pulse; and iii) iterating steps i) and ii), wherein the positions of the seed pulse and the data pulse are incremented along the scan path between iterations, the position of the seed pulse of an earlier iteration being the same as the position of the data pulse of a later iteration, resulting in a birefringent voxel being formed at that position.

[0271] Clause 16. The method of clause 15, wherein an amplitude of the seed pulse is greater than an amplitude of the data pulse.

[0272] Clause 17. The method of clause 15 or clause 16, wherein the substrate is a transparent substrate.

[0273] Clause 18. The method of any of clauses 15-17, wherein generating the seed pulse and the data pulse comprises: generating a source laser pulse, and splitting the source laser pulse using a beam splitter.

[0274] Clause 19. The method of clause 18, wherein the beam splitter comprises: a retardation plate; a first polarization-selective diffractive optic downstream of the retardation plate along an optical path; and a second polarization-selective diffractive optic downstream of the first polarization-selective diffractive optic along the optical path.

[0275] Clause 20. The method of clause 19, wherein the retardation plate is a quarter- wave plate.

[0276] Clause 21. The method of clause 19 or clause 20, wherein the first polarization- selective diffractive optic and the second polarization-selective diffractive optic are each independently selected from a polarization grating, a polarization hologram, and a polarization lens.

[0277] Clause 22. The method of any of clauses 15-21, wherein the positions of the seed pulse and the data pulse are incremented by a non-zero integer divisor of the predetermined distance between iterations.

[0278] Clause 23. The method of clause 22, wherein the non-zero integer divisor is at least 2.

[0279] Clause 24. The method of any of clauses 15-23, wherein the predetermined distance is greater than or equal to twice a wavelength of the seed pulse and the data pulse.

[0280] Clause 25. The method of any of clauses 15-24, further comprising: modulating a polarization of the data pulse.

[0281] Clause 26. The method of any of clauses 15-25, wherein the seed pulse and the data pulse share a common path through a scanning optic and a delivery optic.

[0282] Clause 27. The method of any one of clauses 15 to 26, wherein the iterative steps i) and ii) result in the formation of a plurality of birefringent voxels representing fiducial markers.

[0283] Clause 28. The method of any one of clauses 15 to 27, wherein two or more instances of steps i) and ii) are performed simultaneously in parallel.

[0284] Clause 29. An optical data storage medium obtainable by the method of any preceding clause.

[0285] Clause 30. The optical data storage medium of clause 29, comprising: a transparent substrate; and birefringent voxels embedded in the transparent substrate; wherein the birefringent voxels are in the form of elliptical voids having a semi-major axis, wherein the diameter is in the range of 1 nm to 200 nm.

[0286] Clause 31. The optical data storage medium of clause 30, wherein the transparent substrate is a glass substrate.

[0287] Clause 32. The optical data storage medium of clause 30 or clause 31, comprising a plurality of birefringent voxels.

[0288] Clause 33. The optical data storage medium of any one of clauses 30 to 32, comprising a plurality of birefringent voxels arranged as fiducial markers.

[0289] Clause 34. The transparent optical data storage medium of clause 33, wherein the fiducial markers are two-dimensional fiducial markers.

[0290] Clause 35. A laser writing system, comprising: a laser source for emitting laser pulses; and a write path for receiving input laser pulses from the laser source and focusing seed pulses and data pulses at respective locations in a substrate; wherein the write path comprises: a beamsplitter for splitting said input pulses into seed pulses and data pulses, said seed pulses and said data pulses being spatially separated, an amplitude of the seed pulses being different from an amplitude of the data pulses; and an objective lens for focusing the seed pulses and the data pulses at different respective locations in a transparent substrate, the objective lens being downstream of the beamsplitter on the write path.

[0291] Clause 36. The laser writing system of clause 35, wherein the beamsplitter is configured to split the input pulse into a seed pulse and a data pulse, the seed pulse having an amplitude greater than an amplitude of the data pulse.

[0292] Clause 37. The laser writing system of clause 35 or clause 36, further comprising scanning optics for changing a position of the seed pulse and the data pulse.

[0293] Clause 38. The laser writing system of any one of clauses 35 to 37, wherein the beamsplitter comprises: a retardation plate; a first polarization-selective diffractive optic downstream of the quarter- waveplate along an optical path; and a second polarization-selective diffractive optic downstream of the first polarization- selective diffractive optic along the optical path.

[0294] Clause 39. The laser writing system of clause 38, wherein the retardation plate is a quarter- waveplate.

[0295] Clause 40. The laser writing system of clause 38 or clause 39, wherein the first and second polarization-selective diffractive optics are each independently selected from a polarization grating, a polarization hologram, and a polarization lens.

[0296] Clause 41. The laser writing system of any one of clauses 38 to 40, wherein the first and second polarization-selective diffractive optics are rotatable relative to one another.

[0297] Clause 42. The laser writing system of any one of clauses 38 to 41, further comprising a polarization modulator for modulating a polarization of the data pulse.

[0298] Clause 43. The write head of any one of clauses 38 to 42, comprising two or more write paths and a beamsplitter arranged between a laser source and the two or more write paths, the beamsplitter configured to split a laser pulse to generate an input pulse for each write path.

[0299] Clause 44. A beamsplitter comprising: a retardation plate; a first polarization-selective diffractive optic downstream of the retardation plate along an optical path; and a second polarization-selective diffractive optic downstream of the first polarization- selective diffractive optic along the optical path.

[0300] Clause 45. The beamsplitter of clause 44, wherein the retardation plate is a quarter- waveplate.

[0301] Clause 46. The beamsplitter of clause 44 or clause 45, wherein the first polarization- selective diffractive optic and the second polarization-selective diffractive optic are each independently selected from a polarization grating, a polarization hologram, and a polarization lens.

[0302] Clause 47. The beamsplitter of any one of clauses 44-46, configured to allow relative rotation of the retardation plate, the first polarization-selective diffractive optic, and the second polarization-selective diffractive optic.

[0303] Clause 48. A laser writing system, comprising: a pulsed laser source; a beamsplitter arranged downstream of the pulsed laser source along an optical path, the beamsplitter comprising: a retardation plate; a first polarization-selective diffractive optic downstream of the quarter- waveplate along the optical path; and a second polarization-selective diffractive optic downstream of the first polarization- selective diffractive optic along the optical path; and the laser writing system further comprising focusing optics downstream of the beamsplitter along the optical path.

[0304] 49. The laser writing system of clause 48, wherein the first polarization-selective diffractive optic and the second polarization-selective diffractive optic are each independently selected from a polarization grating, a polarization hologram, and a polarization lens.

[0305] 50. The laser writing system of clause 48 or clause 49, wherein at least one of the retardation plate, the first polarization-selective diffractive optic, and the second polarization- selective diffractive optic is rotatably mounted.

[0306] Other variations or uses of the disclosed technology can become apparent from the disclosure provided herein. The scope of the disclosure is not limited to the embodiments described herein but only by the claims that follow.

Claims

1. A method comprising: Simultaneously generate (201) a first seed pulse (S0) and a first data pulse (D0), the first seed pulse (S0) and the first data pulse (D0) being spatially separated laser pulses, and the amplitude of the first seed pulse (S0) being different from the amplitude of the first data pulse (D0); The first seed pulse (S0) is focused at the first seed position (340(0)) in the substrate (320), and the first data pulse (D0) is focused at the first data position (330(0)) in the substrate (320). The first seed position (340(0)) and the first data position (330(0)) are along the scan path (V). f Separate predetermined distance (jp) v ), in the scan path (V f In the first seed position (340(0)), the first seed position (340(0)) is before the first data position (340(0)); After focusing the first seed pulse (S0) at the first seed position (340(0)) and the first data pulse (D0) at the first data position (330(0)), a second seed pulse (S2) and a second data pulse (D2) are generated simultaneously. The second seed pulse (S2) and the second data pulse (D2) are spatially separated laser pulses, and the amplitude of the second seed pulse (S2) is different from the amplitude of the second data pulse (D2). as well as The second seed pulse (S2) is focused at the second seed position (340(2)) in the substrate, and the second data pulse (D2) is focused at the second data position in the substrate, the second seed position (340(2)) and the second data position are along the scan path (V f Separate the predetermined distance (jp) v ), in the scan path (V f In the second seed position (340(2)), the second seed position (340(0)) is before the first seed position (340(0)) and the second data position, and the second data position is the same as the first seed position (340(0)), resulting in the formation of a birefringent voxel (350(0)). Along the distance of the scan path, in which the second seed position precedes the first seed position and the second data position, and wherein the second data position is the same as the first seed position, resulting in the formation of a birefringent voxel.

2. The method according to claim 1, wherein the amplitude of the first seed pulse (S0) is greater than the amplitude of the first data pulse (D0).

3. The method according to claim 1 or claim 2, wherein the substrate (320) is a transparent substrate.

4. The method according to any one of the preceding claims, wherein generating the first seed pulse (S0) and the first data pulse (D0) comprises: A source laser pulse is generated and split using a beam splitter (644).

5. The method of claim 4, wherein the beam splitter comprises: Delay board (710); A first polarization-selective diffraction optical device (720) is located downstream of the delay plate (710) along the optical path; as well as The second polarization-selective diffraction optical device (730) is downstream of the first polarization-selective diffraction optical device (720) along the optical path.

6. The method according to claim 5, wherein the delay plate (710) is a quarter-wave plate.

7. The method according to claim 5 or claim 6, wherein the first polarization-selective diffraction optical device (720) comprises one of the following: a polarization grating, a polarization hologram, or a polarization lens; and wherein the second polarization-selective diffraction optical device (730) comprises one of the following: a polarization grating, a polarization hologram, or a polarization lens.

8. The method according to any one of the preceding claims further comprises: After focusing the first seed pulse (S0) at the first seed position (340(0)) and the first data pulse (D0) at the first data position (330(0)), and before focusing the second seed pulse (S2) at the second seed position (340(2)) and the second data pulse at the second data position, Perform at least one of the following iterations: Simultaneously, intermediate seed pulses and intermediate data pulses are generated, wherein the intermediate seed pulses and the intermediate data pulses are spatially separated laser pulses with different amplitudes; The intermediate seed pulse and the intermediate data pulse are focused at the intermediate seed position and the intermediate data position in the substrate, respectively, with the intermediate seed position and the intermediate data position separated by a predetermined distance (jp). v Furthermore, in the scan path, the intermediate seed position precedes the intermediate data position.

9. The method according to claim 8, wherein adjacent seed positions (330(0), 330(1)) are determined by the predetermined distance (jp) v Separate the non-zero integer divisors.

10. The method of claim 9, wherein the non-zero integer divisor is at least 2.

11. The method according to any one of the preceding claims, wherein the predetermined distance (jp) v It is greater than or equal to twice the wavelength of the first seed pulse and the first data pulse.

12. The method according to any one of the preceding claims further comprises: The polarization of the first data pulse (D0) and the second data pulse (D2) is modulated.

13. The method according to any one of the preceding claims, wherein the first seed pulse (S0) and the first data pulse (D0) share a common path through the scanning optics and the transmission optics.

14. The method according to any one of the preceding claims further comprises: While generating the first seed pulse (S0) and the first data pulse (D0), a third seed pulse and a third data pulse are generated. The third seed pulse and the third data pulse are spatially separated laser pulses, and the amplitude of the third seed pulse is different from the amplitude of the third data pulse. While focusing the first seed pulse (S0) at the first seed position and the first data pulse (D0) at the first data position, the third seed pulse is focused at the third seed position in the substrate and the third seed position in the substrate, the third seed position and the third data position being separated by a predetermined distance along another scan path, wherein the third seed position precedes the third data position on the other scan path.

15. An optical data storage medium (100) that can be obtained by any of the preceding claims.

16. The optical data storage medium (100) according to claim 15, comprising: Transparent substrate (110); as well as Birefringent voxel (120) embedded in the transparent substrate (110). The birefringent voxel (120) is in the form of an elliptical void with a semi-major axis, wherein the diameter is in the range of 1 nm to 200 nm.

17. The optical data storage medium according to claim 15 or claim 16, comprising a multilayer birefringent voxel (120) and / or comprising a plurality of birefringent voxels (120) arranged as reference markers.

18. A laser writing system, comprising: Pulsed laser source (610); A beam splitter (644) is disposed downstream of the pulsed laser source (610) along the optical path, the beam splitter (644) comprising: Delay board (710); A first polarization-selective diffraction optical device (720) is located downstream of the delay plate (710) along the optical path; and The second polarization-selective diffraction optical device (730) is downstream of the first polarization-selective diffraction optical device (720) along the optical path; and The laser writing system (600) also includes a focusing optics (650) downstream of the beam splitter (644) in the optical path.

19. The laser writing system according to claim 18, wherein the first polarization-selective diffraction optical device (720) comprises one of the following: a polarization grating, a polarization hologram, or a polarization lens; and wherein the second polarization-selective diffraction optical device (730) comprises one of the following: a polarization grating, a polarization hologram, or a polarization lens.

20. The laser writing system according to claim 18 or claim 19, wherein at least one of the delay plate (710), the first polarization-selective diffraction optics (720) and the second polarization-selective diffraction optics (730) is rotatably mounted.

Images