Welcome to the latest issue of PIC International magazine. In issue #7, we begin our technology tour by taking a look at where PICs are today and consider the future requirements for optical chips to serve the world in 2030. Jonathan Marks catches up with Ton Backx, CEO of PhotonDelta, to reflect on this year’s roadmapping activities and find out what the community can look forward to in 2018.

Maintaining our future focus, Mellanox’s Arlon Martin considers how optical solutions could evolve beyond 400G networks, and imec reviews how optical circuits could provide a diagnostic refresh, opening up application areas such as SMART health. So far, so good, but what kind of chip designs will help us get there? Anders Larson, Emanuel Haglund and Johan Gustavsson from Chalmers University of Technology, Göteborg, and Sulakshna Kumari, Gunther Roelkens, and Roel Baets from Ghent University-imec share their thoughts on empowering silicon with vertical cavity lasers to light up the chip.

Turning to other opportunities for PICs, Iñigo Artundo, CEO of VLC Photonics, reviews what it takes to implement 3D optical imaging systems. And reading on through the magazine, you’ll discover how LED manufacturers are helping to bring a plug-and-play approach to Lifi, which could help in making the technology a feature of future mobile devices.

Lastly, we’re pleased to share highlights of the 2017 EPIC Workshop on photonic integrated circuits, which was hosted and sponsored by OITDA, PETRA and the Photonics and Optoelectronics Network PHOENIX+.

Enjoy the issue!

James

“We have capitalized on the extraordinary level of trust that was built up in Brabant”, Ton explains. “Since the meeting in June, we've been working out the details gleaned from the 6 parallel technical working groups that were formed. Based on that wealth of data and ideas, we've decided to redefine the structure a little bit, so that we can streamline the interactions, make the findings more coherent, and decide what is still missing. 

We're doing this first report together with Berenschot, who are well-known technical strategy consultants in this part of the Netherlands. Together, we've been learning from the experiences of our colleagues at AIM Photonics Academy who set up the IPSR International. At the same time, we’ve been making sure that we're setting up a collaborative process that is manageable and yet has short communications lines.” 

“Berenschot has been in touch with each of the six working groups and has been filling in the results according to the structure we agreed upon. They are nearly finished with what is a very complex task. I hope that by the end of 2017 we will be able to release the first draft of the Global Roadmap. It will then be opened for on-line discussion amongst those who attended the first World Technology Mapping Forum and any other parties who believe they can contribute. Of course, I don't expect it to be complete because our discussions in June this year were simply a starting point. But it a good start for the next WTMF-2 which is coming up June 20-22nd 2018 in Enschede, the Netherlands” 

Start of an important process

“To keep discussions going, we also have fortnightly collaborative meetings with our colleagues at AIM Photonics and the IPSR. They have their established roadmapping procedure centered around silicon-based photonics and we actively participated in their fall AIM Photonics meeting in October to bring in the Indium Phosphide expertise”. 

“I believe we need to fine-tune the structure of the conversation to ensure that not-only research goals are covered well from a technology perspective, but also that everyone has a clear understanding of industry needs and expectations. In that way, the roadmap will show what kind of functionality can be expected at what point in time to produce new products. We also need to forecast what sort of cost levels can be expected, so we can check whether these market propositions are feasible and look for tipping points. Our roadmap needs to support both these lines of thinking”. 

“Integrated Photonics has developed using various materials for the wafers. In the Netherlands the expertise is centered around indium phosphide and silicon nitride. Other countries like Belgium, France, the USA, and Taiwan have a different focus on silicon photonics, which is the material used in the existing micro-electronics industry. Several different hybrid solutions are being developed now for next generation applications, which will mix different materials. That has several advantages, but will also raise all kinds of new connection issues between the two”. 

Investors and industry interest

Many parties are closely watching the outcome of the deliberations from the technology forum. Large private and public investments will need to be made to scale up the manufacturing volume of next generation chips. Investors are naturally keen to find, select and back the winning technologies that have a proven product-market fit. 

The photonics manufacturing industry wants to understand the technology choices that have been made to bring next generation products to market at a price that customers will pay for. These companies don't want to be confronted with a range of technological solutions for a product that they want to launch into the market next year. 

Why silicon and indium phosphide need each other

“If we look at next generation devices that will come on the market in 2018/2019, you will see several competing solutions from the separate silicon and indium phosphide worlds. Globally, a lot of work is being focused now on the silicon photonics ecosystem. But one of the drawbacks of silicon is that you can only produce passive circuits. When the circuit complexity is relatively low, you can make use of external discrete light sources or simple light  sources integrated on chip. There are many manufacturers who can supply the discrete lasers. In the short-term there is no problem. But let’s look ahead 5 years.” 

A world of 1600 Gb/s

“By 2023 the global market will be asking for transceivers that operate at a speed of 1.6 Terabit per second. If we take that as an end-goal, then compare it with the situation today where 0.1 Terabit per second is common. Today, communication links between datacentres all use optical links with single mode fibres. But the processing of the data in the data centres is done electronically.” 

“The problem is that when you go from 0.1 to 1.6 Tb/sec, the light part can cope relatively easily, but the speed of the electronics becomes a severe bottleneck. With current technology, you can find commercial electronics modules operating at speed of 10 Gb/second. To get a transceiver operating at 100 Gb/sec, you build ten 10 Gb/sec electronic channels and operate them in parallel. The optical side already works at 100 Gb/sec. There is, of course, a conversion process to split the optical signal over the 10 parallel electronics channels.” 

“Suppose we go to a transceiver operating at 400 Gb/sec, then this will require 40 parallel channels on the electronics side if the same electronics is applied. That is a lot more complex to achieve.”

Solutions for the bottleneck 

Electronics is still important in the coming years. But it is also a bottleneck and it affects the way systems are going to be engineered. 

“To come up with 400 Gb/sec transceivers for launch in 2019, some manufacturers apply electronics operating at 25 or even 50 Gb/sec. That would require only 8 parallel channels instead of 40.  But this doesn't solve the problem by the time we've moved to a world of 1600 Gb/sec by 2023.” 

“Having 160 or even 24 electronic channels operating in parallel raises the question where the conversion will be done. If the splitting, processing channel conversion is done on the optical side it can be done very efficiently, but additional signal processing complexity is required at the optical side. CMOS technology based electronics will become significantly more complex when it has to operate at bit rates beyond 25 Gb/s” 

“Many parties working on the 400 Gb/sec solutions with silicon based PICs have concluded that for the moment, the chips will do limited optical processing. They will make use of a single or a few external laser sources to provide light to the silicon transceiver. The systems operating in that way will not have very long-distance communication capabilities. The distance they can bridge is not much more than 2-3 kilometres. While that may sound a lot, next generation distributed data centres require high speed transceivers that can bridge distances of 50-100 kilometres. Indium phosphide transceivers have already bridged a distance of 50 kilometres and beyond. So it may be expected that next generation transceivers applied for short distance may predominantly be based on silicon based photonics fed by an external laser while indium phosphide based  high speed transceivers will be used for longer distances and short distance.” 

Exponential energy challenges mean disruptive solutions

“Let me share some numbers that came out of the WTMF technical workshop discussions. We expect that the 1.6 Tb/sec transceivers we will be using by then will need to bridge distances of 50-100 kilometres. Datacentres used for cloud functionality are being built by enterprises like Amazon, Google, Apple, Facebook, and Alibaba. They approximately have to double the capacity every year.” 

Google Eemshaven in 2016

“You may know that Google has built a €600 million datacentre in Eemshaven in the North of the Netherlands. They report they have laid 16,000 kilometres in cabling. Now, although the energy being supplied to the new facility is currently 100% from renewables, the power consumption is around 250 Megawatts. The power station in the area only has a capacity of 1.2 GW available. But at the current rate of growth using conventional technology, the demand will be around 500 Megawatts by the end of 2018, and close to 1.0 GW by the end of 2019. That puts a very difficult demand on local power resources. Doubling the generation capacity of that power station is going to take seven years, so that is not going to solve the issue. So, you will have to split up that datacentre capacity, so it can tap into other energy resources from a region nearby.”

The impact of photonics

“The good news is that the introduction of photonics into datacentres will impact their power consumption quite considerably. We estimate that instead of doubling each year, power consumption will go down to a factor of around 1.3 - 1.7. But it will still require more datacentres operating in parallel to satisfy the exponential global demand for bandwidth and processing power.”

“There is another trend we need to consider. As the complexity of these next generation chips increases, light losses in passive light processing need to be compensated for using light amplification and additional light sources. That puts limits on the chip's functionality ,if only passive circuits can be used, because at a certain moment, all the light is gone. Active laser sources on the chip will be needed to regenerate and amplify the light. We believe that indium phosphide is going to be needed in all the scenarios because when you need hundreds of laser sources or light amplifiers, you are not going to be able to use external discrete lasers anymore. By 2021/2022 we need to have a clear structured plan of how industry is going to do this, so roll out of the second chip generation is ready by 2023.”

“This realization that the current technology cannot keep pace that motivated me to start travelling round the world to urge others in photonics to collaborate and solve this pressing problem. I'm pleased to say that by openly collaborating we are getting much closer to a solution. Our colleagues at AIM Photonics Academy in Boston are doing some excellent work. And in Europe there are around a dozen hotspots that have put photonics near the top of their high-tech agenda. You will hear more about the European Photonics Alliance next year. PhotonDelta and AIM Photonics Academy are already convinced that working together is going to be essential to solve the packaging issues and to build very complex chip structures with active circuits on board. Joint research and knowledge exchanges are being planned for the middle of 2018 because there is no time to waste.” 

Moore's Law is no longer fast enough for photonics 

PhotonDelta recently announced the launch of a national  Photonic Integration Technology Centre (PITC) as part of a new Dutch National Photonics Agenda. 

“This is the result of three photonics “centres of excellence” working together, in the areas around Eindhoven, Enschede and Nijmegen. Each region has complementary micro-electronics knowledge and PIC expertise. And by global standards we're physically very close to each other. The PITC will offer its services through new alliances currently being established all over the world. It's a global organization with a Dutch accent, encouraging Open Innovation development.”

“Manufacturers, integrators and end users all have the same goals. They want reliable, tested, stable systems at the lowest possible cost. PhotonDelta's PITC will play a leading role helping companies accelerate in these very important packaging and reliability-engineering phases to retain and grow volume manufacture of photonics devices in Europe. The PITC will support the chip fabs like SMART photonics, Heinrich Hertz and LioniX International to ensure that their second-generation chip production technology can scale at exactly the moment there is the right product-market fit. The PITC will also do the further development in materials processing and systems development to support innovative enterprises like Technobis and EFFECT Photonics.”

PhotonDelta Cooperative established

“We have also established a separate PhotonDelta Cooperative as a new independent trusted, legal entity. Research organizations and industry are being encouraged to join to get early access to fundamental and applied research into next generations chips, circuits and systems and the associated IP. This is essential to ensure that we maintain and grow the high-pace of research that is needed now, but also that we keep listening to understand what the market needs tomorrow. The party that pays gets ownership of the IP, but the IP generated in that Open Innovation structure must be made available license-free for further research. If we get bogged down in endless IP ownership discussions pace of development will be too low whereas speed of development is critical in supporting the market. ”

“We have learned from our Canadian and Irish colleagues about the need to encourage cross-fertilization between research, development and industry. We need to encourage people to build a career across these sectors, in the same way this was initiated  in microelectronics in the 1970's. So, you might start in fundamental research. Once some of the researchers have validated a technology they can move on to application development within an organization like PITC. Later they would go to work within industry to ensure the faster execution of that technology. With product life cycles so short these days, we think this is the best way to speed up knowledge transfer. If we keep a simple and transparent structure, then it is easy to make sure it works!"

Q1 - What makes silicon photonics so appealing?

Cloud datacentres are looking for technology sweet spots that can enable them to scale their networks to next-generation speeds at lower cost, with lower power consumption, and with faster manufacturing ramps than previous generations. Silicon photonics addresses this sweet spot. Silicon photonics chips are already widely used in today’s 100Gb/s networks with four 25Gb/s lanes, supporting datacentres with reaches of 500 metres to 2 km. For the next generation of 400Gb/s networks, Mellanox has already demonstrated the capability to transmit at speeds of 100Gb/s per lane -- four times faster than today’s products -- using the same size chip. This will lower networking cost and improve density. In addition, the energy efficiency improves by roughly a factor of two, meaning that datacentres run cooler and cost less to operate. Because silicon photonics chips eliminate the assembly complexity of traditional optical transceivers with hundreds of lenses, filters, isolators and other optical components, the manufacturing is far simpler. While traditional optical assembly processes can take several years to scale to high volume, silicon photonics optical engines are fabricated in standard foundries, so the volume ramp-up can be extremely fast. For hyperscale datacentre networks of 500m to 2km in reach, silicon photonics really hits the sweet spot.

Q2 - What are the key roles for optical chips in Mellanox's portfolio of solutions? 

Mellanox is well-known as a leading supplier of end-to-end Ethernet and InfiniBand intelligent interconnect solutions and services for servers, storage, and hyperconverged infrastructure. Recently, at Supercomputing 17, Mellanox announced what we believe is the world’s most scalable switch platform based on HDR 200G InfiniBand Quantum switch technology. With up to 1600-ports in a single platform, Quantum Switch Systems enable the highest performance while reducing datacentre network expenses by 4X. The new Quantum switch will use the same QSFP pluggable form factor as with previous generations, but double the speed to 200Gb/s. Silicon photonics transceivers easily fit in this compact package, and support the higher bandwidth and the longer reaches required by web 2.0 and cloud data centres. Not all 200Gb/s networks perform the same. Customers want the fastest switching times, the lowest latency and the lowest Bit Error Rate (BER). Our Silicon photonics-based optical transceivers ensure the highest throughput with the lowest BER for any InfiniBand or cloud data centre reach. Our transceivers have a BER of 1E-15 which, at 200Gb/s transmission, is one bit error every hour or two. Many competing solutions have a BER of 1E-12, which is 1000x worse, with an error occurring every five seconds. Silicon photonics transceivers just perform better, in my opinion. 

Q3 - Looking ahead at even greater utility of CPU and storage elements in the cloud - what are the pinch points and how do you see optical solutions evolving?

For the past 20 years, datacentre networks have been built around front-panel pluggable transceivers. The transceiver speeds have scaled with the switching speeds of the switches: 1 Gb/s transceivers for 1Gb/s ports; 10Gb/s transceivers for 10Gb/s ports; etc. Having ports on the front panel, allows many options for technologies in their networks. For example –

• Direct Attach Copper (DACs) could be used to connect servers to Top-Of-Rack switches
  
• Active Optical Cables might be used for storage and appliances up to 100m away
  
• VCSEL optical transceivers for reaches of several hundred meters on multi-mode fibre
  
• Silicon photonics optical transceivers for reaches of up to 2km or more on single mode fibre
  

But, sometime after the 400G generation of networks, it’s likely that the “Front panel pluggable” model breaks and that the optics will migrate inside the switch itself, either on the board with the switching chip or co-packaged in a Multi-Chip Module (MCM) with the switching chips. Mellanox is an active participant in the Consortium for On Board Optics (COBO), an industry effort to standardize packaging for on board optics. For COBO and MCM applications, the advantages of silicon photonics integration become large. SiP chips can work at higher temperatures than traditional laser technologies. Smaller sizes mean that SiP solutions can be closer to the switching chip. Functional blocks, like the modulator driver, can be incorporated into the switching ASIC itself. Finally, the entire switch assembly process favours the electronics-style assembly of silicon photonics.

What is biophotonics-on-chip?

Chances are, you use photonics every day without even realizing it: glass fibres enable you to use a computer or watch television without any problems. With the help of light, these glass fibres send data much more quickly and power efficient than with traditional digital cables. 

You can also do the same thing on a chip. Using ultra-small ‘fibres’ and structures, you can direct light on to a chip and carry out a whole range of tasks. Those tasks can involve processing or sending data, but biological tasks are also possible. In fact, light is the most frequently used medium in medical diagnostics – just think of microscopes and spectroscopes. Light enables you to count or visualize cells, measure the properties of materials and tissue, define a DNA sequence, etc. 

Small ‘fibres’ of silicon nitride (SiN) are produced on top of a silicon chip. These waveguides direct the light along a well-defined path over the chip and along detection sites.

Biophotonics-on-chip is a fairly recent area of research that is becoming very important in the medical sector for diagnosis, treatment and follow-up. Doctors will soon be able to use the technology to analyze a blood sample without bulky (fluorescence) microscopes, as well as examine a tissue sample without large spectroscopes. 

It is quite a challenge to make photonic structures very small and then combine them into a photonic circuit capable of carrying out a specific task with great efficiency and reliability. If you produce structures using silicon, such as computer chips, or based on silicon-compatible materials (SiN), you can incorporate electronic and photonic functions, creating a smart and compact system. And if you can do that, you can easily make hundreds and thousands of systems function alongside each other at the same time, which means you’ll also get the result much faster than with a single system on its own. Thanks to biophotonics-on-chip we will soon have small, cheap test chips that will assist doctors in their decisions. 

Some of the photonic components made by imec: spectrometer, fibre waveguide, waveguide and multi-mode interferometer.

Counting and viewing cells

Fluorescent labels are often used to analyze a blood sample in the traditional way. These labels are in fact molecules that bind specifically with, for example, (parts of) a bacterium, gene, cancer cell, etc. Using a cytometer, the blood sample and hence also the labels are illuminated, making them radiate fluorescent light that can be detected. This enables the number of bacteria and cancer cells to be counted or the presence of a specific gene or DNA sample to be determined. 

A recent development in this area is a photonic structure called a ‘focusing grating coupler’. A grating coupler is usually used to couple the light coming from a laser (which shines on the chip) in the waveguide paths on the chip and to shine the light coming out of the waveguides back from the chip (for example, to a detector that is not integrated on a chip). In the new development, the focusing grating coupler allows light, moving in a waveguide, to shine outward from the surface, to create an upward beam of light on top of the chip. 

Focusing grating couplers can send the light in waveguides outward from the surface, enabling them to efficiently illuminate the cells in the microfluidic channel located above.

The microfluidic channel that the cells flow along runs through this light beam. In this way, the cells with fluorescent labels are illuminated, after which they emit fluorescent light. This fluorescent light is captured by the ‘diffraction gratings’, which sort the light by wavelength. As a result, various fluorescent labels can be detected at the same time. This is a very good example of how compact photonic chips can be used to count cells, even different types of cells (with different fluorescent labels) at the same time. The big advantage of this approach is not so much that the cell can be optically detected, but that hundreds of these structures can work in parallel, having a huge impact on the throughput of these measurements. 

Not only is it of value for counting cells, but it is also good for looking at the morphology of the cells. Here again imec has developed an integrated solution: the lens-free microscope. 

Spectroscopy in miniature

Spectroscopy is used in medicine to detect certain substances in tissue, skin or areas of the brain, such as cholesterol, lactic acid and ethanol. Melamine in milk, phthalate in toys, contamination in meat or the authenticity of medical drugs can also be detected with spectroscopy. The substances are detected by their specific interaction with wavelengths of light. 

There are many forms of spectroscopy, including absorption, reflection, fluorescence and Raman spectroscopy. Imec is aiming to develop a mini-version of the Raman spectrometer on a chip. This would enable a compact little device to be produced for measuring specific substances in a blood sample regularly and non-invasively. This is not possible with existing desktop Raman spectrometers. 

The major challenge in developing a Raman spectrometer is balancing the very small usable signal against the large background signal. That’s why the detector has to be very sensitive. One of the best-known spectrometer designs is the Michelson interferometer. A beam of light is divided into two beams that take different paths before coming together and interfering. This enables tiny differences in the wavelength to be measured. The disadvantage of this design – particularly if you want to miniaturize it – is that two mirrors are used, one of which moves. Unless the moving mirror is in absolutely the correct position, the measurement is incorrect. 

Design of the Raman spectrometer on chip.

Imec has developed a (patented) solution with no moving parts in which hundreds of structures – interferometers – are used next to each other. Light is shone on the tissue and the scattered light is collected by a collimator. This divides the light – with the help of a beam-shaper – across the various interferometers. Each interferometer is a little smaller than the previous one so that tiny differences in wavelengths can also be measured, as is the case with the Michelson interferometer. 

A hypersensitive sensor based on light & sound

Photoacoustics is a fast, relative cheap and harmless way of producing images of the human body. It can be used, for example, in research into skin and breast cancer. The photoacoustic effect was discovered in 1880 by Alexander Graham Bell, the inventor of the telephone. He illuminated a block of selenium, which created a weak sound (hence photo = light and acoustics = sound). In fact, light and sound are both forms of vibration. It’s just that we can’t ‘hear’ light, although it can be converted into sound. 

With photoacoustics, very brief laser pulses are directed at the patient’s body. A different color of light is chosen, depending on the tissue. When one of these pulses touches the tissue, it is converted into heat. The tissue expands and then contracts again, creating a change in pressure, which moves again as ultrasound. This signal can be picked up by a sort of microphone. The ultrasound can be used to gather spectroscopic information about a material, or else it can be converted into an image. The big advantage of photoacoustics is that there is no background signal, which makes it a highly sensitive technique. 

For its photoacoustic sensor-on-chip, imec uses a membrane with integrated waveguide. When the membrane is moved by a sound wave, the waveguide is stretched and this movement can be recorded.

Photoacoustics are already used extensively in medical research, although not yet for diagnosing patients, because the technology is still too expensive. This is where imec aims to introduce a change by making a photoacoustic sensor on chip. One important component for this is the ‘microphone’, which must be able to pick up ultrasound. The ‘mic’ consists of a silicon oxide membrane with an integrated photonic waveguide. When the membrane moves under the pressure of a sound wave, the waveguide is stretched and this movement can be recorded. 

Once it becomes possible to miniaturize spectrometers and photoacoustic sensors, the chip may be integrated in a pen like the one in the drawing. The doctor can then use the pen to scan the patient’s skin looking for disorders, such as skin cancer.

The two leading platforms for the production of PICs are InP and silicon. The primary advantage of the former – predominantly used in transceivers for wavelength division multiplexed telecom systems – is the monolithic integration, on a single chip, of all active and passive elements, including lasers as light sources. In contrast, the latter – commonly referred to as silicon photonics – has the advantage of a far lower propagation loss in the waveguide. With this material system, there is the benefit of the use of CMOS fabrication processes, which are suited to highvolume, low-cost manufacturing.

Arrays of hybrid vertical-cavity lasers with intra-cavity SiN waveguides

Waveguides for silicon photonics are often made from silicon or SiN. Silicon is transparent at wavelengths beyond 1.1 μm, so this technology, which is based on silicon-on-insulator structures, is the most common platform for telecom and datacom transceivers. If transparency in the visible range or near-infrared (below 1.1 μm) range is required, silicon is unsuitable, and SiN waveguides tend to be adopted. SiN technology is often used to fabricate PICs for biophotonics and life sciences, where wavelengths in the visible and very-near-infrared are of particular interest.

Figure 1. A bio-photonic SiN PIC sensor chip with integrated micro-ring resonators coated by receptor layers, a heterogeneously integrated current tuneable laser, and flip-chip integrated laser driver and grating coupled photodetectors.

One major challenges for any PIC made with silicon photonics is the integration of the light source. Difficulties arise from the lack of a direct bandgap for silicon and its compatible compounds, such as SiGe and SiN. A direct bandgap material is essential, as it is a prerequisite for efficient light generation and amplification.

Figure 2. The concept of hybrid vertical-cavity laser integration on SiN PICs. Upper: Cross-sectional view of the ‘half III-V VCSEL’ bonded to a dielectric DBR on silicon with an intra-cavity SiN waveguide with a weak diffraction grating on top. During each round-trip in the vertical cavity, a certain fraction of the photons stored in the cavity is tapped off to the in-plane intra-cavity waveguide. Lower: Top view of the SiN waveguide and grating onto which the ‘half III-V VCSEL’ is bonded.

Most of the producers of silicon PICs select one of three approaches to integrate the lasers to the chips. Their least mature option is monolithic integration, which involves hetero-epitaxial growth of compound semiconductors on to silicon. One alternative is heterogeneous integration, which involves attaching epitaxial compound semiconductor structures to silicon using either die or wafer bonding. The resulting structure is processed to form lasers, allowing the merger of otherwise incompatible materials to yield a high-performance laser. There is also a third approach: hybrid integration. In this case, light from a stand-alone laser is coupled to the on-chip silicon or SiN waveguide.

Figure 3. The GaAs-based surface-emitting version of the hybrid vertical-cavity laser, lacking the intra-cavity SiN waveguide and grating for in-plane emission.

Regardless of the approach, most engineers incorporate an in-plane laser in their PIC. However, its performance is not ideal: it has high bias and modulation currents, it operates with a low power conversion efficiency, and it has a large footprint.

VCSEL virtues

For data communication, sensing and some highpower applications, the most common laser used today is not an in-plane device, but a VCSEL. Its success stems from: its small optical mode and gain volumes, enabling efficient operation and high-speed modulation at low currents; and its vertical geometry, which provides surface emission and enables dense twodimensional arrays and low-cost fabrication and testing.

The success of VCSELs raises an obvious question: can it, or a device like it, provide the light sources for silicon photonics? If it could, it would empower silicon photonics with a class of lasers that offer a low current, a high efficiency and a small footprint.

Given all this promise, it is of no surprise that many groups have tried to develop silicon PICs that feature flip-chip integration of long-wavelength (InP-based) and short-wavelength (GaAs-based) VCSELs over optical coupling elements, such as gratings and mirrors. However, this approach does not deliver a wafer-scale process. Instead, it involves accurate, time-consuming alignment of individual VCSELs in a back-end process.

What’s needed is a wafer-scale compatible process. Fulfilling this goal is our partnership between researchers at Chalmers University of Technology and Ghent University-imec. Together we have developed a process for the heterogeneous integration of hybrid vertical-cavity lasers.

At the heart of our technology is the formation of a hybrid vertical-cavity laser via the bonding of an epitaxial III-V structure – it contains the upper reflector and an active region that provides optical gain under current injection – to a lower reflector on the silicon substrate. This approach forms a hybrid vertical cavity.

With our technology, light is coupled to an in-plane silicon or SiN waveguide with an optical element in the cavity. This element taps off power to the waveguide.

Figure 4. Optical field intensity and refractive index along the optical axis of the hybrid vertical-cavity laser.

Our technology is applicable to many material systems. It is compatible with GaAs, InP, and GaSbbased materials, for emission in the very-near-infrared, near-infrared, and mid-infrared, respectively; and with the development of GaN-based materials, it may become useful for visible light sources.

Figure 5. Illustration of the fabrication process for the GaAs-based hybrid vertical-cavity laser. (1) Bonding of III-V die on the silicon part using adhesive bonding. (2) Removal of the GaAs substrate. (3) Top contact metallization, mesa etching, selective oxidation, and bottom contact metallization. (4) Planarization with BCB and deposition of pad metals.

We are not the only team pursuing this type of approach. However, we distinguish ourselves by working with GaAs-based materials and targeting light source integration for the very-near-infrared. Meanwhile our peers, the group of Connie Chang-Hasnain at the University of California at Berkeley and the group of Il-Sug Chung at the Technical University of Denmark, use InP-based materials as near-infrared sources.

The GaAs-based, hybrid vertical-cavity laser technology that we are developing forms part of the European Horizon 2020 project PIX4life. The project, co-ordinated by imec and involving 15 partners across Europe, aims to establish a SiN PIC pilot-line for life science applications in the visible and very-near infrared. An intended outcome of our efforts is a line that will aid product development for a broad range of industrial customers.

One example of a device developed on this line is a bio-photonic sensor (see Figure 1). This chip contains micro-ring resonators coated by receptors. The receptors selectively bind target analytes, leading to a shift in the micro-ring resonance frequency. Sensing these shifts is an on-chip current tuneable laser, which interrogates multiple sensors with different receptors. For this kind of sensing PIC, the hybrid vertical-cavity laser is potentially the ideal light source.

To enable on-chip integration of hybrid vertical-cavity lasers, we are developing an integration platform that features a dielectric distributed Bragg reflector (DBR) buried under the waveguide on the silicon substrate (see Figure 2). We accomplish this by bonding the epitaxial structure with the upper DBR and the active region to the silicon wafer to form the hybrid verticalcavity. A shallow grating etched in the intra-cavity waveguide diffracts light to the in-plane waveguide. Note that an additional benefit of the buried DBR is that it can improve the efficiency of grating couplers used to couple light from the PIC into the likes of optical fibres or flip-chip integrated detectors.

Surface-emitting lasers

An ultimate goal of our project is to demonstrate on-chip waveguide-coupled integration of hybrid vertical-cavity lasers. However, we begin by taking a step towards this: a surface-emitting version, which is a VCSEL. By taking this route, we can develop and implement the integration architecture, and then investigate its performance characteristics and limitations.

Our intermediate structure features a SiO₂/Ta₂O₅ DBR on silicon for the lower reflector, on which we bond a GaAs-based epitaxial structure with a p-type AlGaAs DBR, an active region with strained InGaAs quantum wells, and an intra-cavity n-type AlGaAs contact layer (see Figure 3). As is often the case for GaAsbased VCSELs, we use selective oxidation to form an oxide aperture for transverse current and optical confinement.

We have designed our hybrid vertical-cavity laser for 850 nm emission. With the design we are pursuing, the optical field extends over the GaAs-based and silicon-based parts of the hybrid cavity (see Figure 4).

Fabrication of our developmental structure begins with the preparation of its two parts (see Figure 5). They are formed by depositing the dielectric DBR on the silicon wafer and growing the epitaxial structure on the GaAs substrate by MOCVD. An additional thin layer of SiO₂ is deposited on the dielectric DBR, followed by spin-coating a film of the polymer DVS-BCB, which acts as an agent for adhesive bonding.

Figure 6. Upper: Optical microscope images of a single hybrid vertical-cavity laser and an array of such lasers. Lower: Scanning electron microscope images of a focused ion beam cross-section through the hybrid vertical-cavity laser (before BCB planarization).

Basic performance characteristics are obtained by measuring the optical output power and voltage as a function of current at temperatures up to 100°C. Plots show that performance, evaluated in terms of the temperature-dependent threshold current and slope efficiency (see Figure 7(a)), is similar to that of an ordinary oxide-confined VCSEL. This demonstrates the potential of our approach to integration.

Figure 7. Left: Output power and voltage as a function of drive current at temperatures from 20°C to 100°C for an 850 nm hybrid vertical-cavity laser with an oxide aperture diameter of 10 μm. Right: Dependence of threshold current on temperature.

Basic performance characteristics are obtained by measuring the optical output power and voltage as a function of current at temperatures up to 100°C. Plots show that performance, evaluated in terms of the temperature-dependent threshold current and slope efficiency (see Figure 7(a)), is similar to that of an ordinary oxide-confined VCSEL. This demonstrates the potential of our approach to integration.

Figure 8. Left: Output power and voltage as a function of drive current at 25°C to 85°C for an 850 nm hybrid vertical-cavity laser with an oxide aperture diameter of 5 μm. Right: Results from data transmission experiments at 10 Gbit/s up to 85°C and 25 Gbit/s at 25°C.

We have found that if an appropriate thickness is used for the bonding interface, the threshold current is only weakly dependent on temperature (see Figure 7(b)). However, the output power saturates at relatively low currents, due to the high thermal impedance - it stems from the low thermal conductivity of the dielectric DBR. One way to address this is to integrate metallic heat spreaders or thermal shunts. These modifications increase the efficiency that heat is conducted to the silicon substrate, leading to reduced thermal impedance and ultimately a higher output power.

Another promising application for our shortwavelength, hybrid vertical-cavity laser is as the light source in integrated transmitters for wavelength division multiplexed optical interconnects, where the SiN PIC is used for multiplexing. To evaluate its potential, we have studied the dynamics, measuring the modulation bandwidth and data transmission rates. We found that we could transmit data at up to 25 Gbit/s at room temperature and 10 Gbit/s at 85°C by using a smaller aperture and a modulation bandwidth exceeding 10 GHz (see Figure 8).

Our results showcase the potential of our hybrid vertical-cavity laser for the integration of low-current, high-efficiency, small-footprint light sources on silicon PICs. However, there is still work to do. We must demonstrate that a similar performance is possible with in-plane emission, by incorporating an intracavity waveguide and diffraction grating. The good news is that the signs are promising, with simulations suggesting that the lower waveguide/grating/DBR combination can be designed to pin the polarization of the cavity mode in the direction needed for controlled, efficient coupling to the waveguide.

3D imaging systems are based on the measurement of how far away each point in a 3D space is from the emitting device, as well as the direction to that point. The combination allows for the creation of a full 3D model of the surrounding environment. LIDAR emerged in the 1970's as a ranging and 3D imaging technology that uses optical signals in visible (400-750 nm approximately) and Near InfraRed (NIR, 750-1400 nm) spectra. Such light signals, after hitting the objects around them, are absorbed, scattered out, or reflected back to the emitter. As the used wavelengths are ~100.000 times smaller than radio signals commonly used in RADARs, LIDARs can offer better resolution and depth precision.

LIDAR example: 3D imaging of a house (image - University of Washington).

Some examples of different applications making use of LIDAR nowadays are:

• Intelligent vehicles (pedestrian protection, adaptive cruise control, autonomous driving, traffic monitoring, etc.),
  
• Robotics (handling of complex or small objects, collision avoidance, etc.),
  
• High-resolution short-range mapping (for example, to be applied in 3D printing), or long-range for terrain and ocean cartography,
  
• Drone navigation (flight safety, tracking, search and surveillance, etc.),
  
• Medical (corneal imaging for contact lens fitting, 3D vision for robotic microsurgery, OCT, etc.),
  
• Atmospheric sensing measurements (detection of dust and gas pollutants in air, ground water, subsurface soil, weather avoidance, airflow characterization, etc.), 
  
• Consumer electronics (augmented reality, gaming systems, etc.).
  

Today, commercially available high-end LIDAR systems can range from $1,000 to upwards of $70,000, which can limit their applications where cost must be minimized. Most LIDAR systems are still very bulky, not robust and expensive, and therefore not suitable for mass introduction in certain markets. 

Implementation options

While lower-cost mapping devices use stereoscopic imaging, most LIDAR systems -- for example, autonomous cars -- are based on Time of Flight (TOF) techniques that measure the round-trip delay of a light pulse to the target. This type of LIDAR consists of a light source (for example, a laser), a detector, and some electronic timing circuitry. TOF uses the differential time between the transmission of a high-power pulse and the reception of an echo to provide range. The simplicity of the optical parts in a pulsed LIDAR makes it attractive for many applications. However, to achieve a precision below ~10 μm, the electronic circuits must be able to measure the round-trip delay of the light pulse with an accuracy of approximately 70 fs, which is very challenging for current electronics.

Uber car with bulk LIDAR system fitted on its roof (image - Wikimedia Commons)

As an alternative to TOF LIDAR, a continuous wave system can use the same local oscillator reference for transmit and receive. A Frequency-Modulated Continuous-Wave (FMCW) system radiates continuous transmission power and, at the same time, the operational frequency is swept. As the transmission signal is modulated with a chirp pattern, the distance is accomplished by comparing the frequency of the received signal to a reference. Range resolution for FMWC depends on how high is the chirp period and its frequency range. Moreover, while TOF requires the use of the Doppler effect to measure velocity, FMCW has by itself the ability to calculate the relative velocity. FMCW can thus relax the required precision in the round-trip delay measurement in exchange for a more sophisticated optical architecture, requiring an accurate modulation of the laser frequency. 

In both cases, the laser beam steering can be done mechanically by servo systems, which work with rotatory motors or micro-mirrors that may not be very stable and somehow limited in speed, or Optical Phased Array (OPA) antennas, which consist of an array of several emitters that, by aligning their phases, interfere constructively in the far field at certain angles. 

The attraction of OPAs

OPAs have gained interest in recent years as an alternative to traditional mechanical beam steering because they lack inertia, which limits the ability to reach a large steering range at a high-speed, and can be implemented reliably in a photonic circuit by integrating the steering elements with an on-chip laser for very high-speed operation. OPAs can be fabricated in commercial CMOS foundries, much like electronic phased arrays have been for decades, with the advantages of smaller size and weight, high reliability and yield, and overall system scalability, not only in complexity and performance, but also in production volumes.

Ideal characteristics required for such beamforming OPAs are broad instantaneous bandwidth, continuous amplitude and delay tunability and, capability for feeding large arrays. Photonic beamforming can be implemented with true time delays, with the challenge of the high propagation losses required, or by shifting the optical phase, with the sensitivity issue of the optical phase and its susceptibility to thermal drift. Long-range power transmission for LIDAR systems usually require output beams to have a small diffraction angle and high power. Both can be addressed with a large aperture size and a power distribution network

Photonic integration examples

Silicon photonics has been proposed for many years [1] as a platform for implementing OPAs, and here the beam could be steered to an angle with heater electrodes or vertical grating couplers. Wavelength properties can also be used for steering the beam when using, for example, out-of-plane vertical grating couplers because the emission angle of a grating is dependent on the wavelength. It is relatively easy to increase the 3 dB field-of-view (FOV) by reducing the width of the grating coupler waveguides. Here, a trade-off must be made between the FOV and the beam width, as the number of addressable spots in free-space depends on the number of radiating elements of the OPA.  

Some of the widest steering range (80º) and smallest beam divergence (0.14º) has been demonstrated using silicon photonics. A combination of non-uniform emitter spacing was employed to suppress lobes that limit the steering range, weak gratings, wide-angle emitters and a large collection of phase-controlled emitters, allowing beam-steering over 500 resolvable points [2].

However, nonlinearity-induced phase changes and material loss limit the scalability and power output of silicon photonics arrays. Silicon nitride, while mostly a passive platform, has been proposed for its high-power capability and robustness to fabrication induced phase variation, and there are already demonstrations of a 1024 antenna phased array operating at a wavelength of 1550 nm, with an overall 4×4 mm2 aperture size [3]. A main beam output power of 400 mW was achieved, being near diffraction limited with a FWHM spot size of 0.021º×0.021º. Moreover, using the same silicon nitride fabrication process and design architecture, the first large-scale visible phased array at 635 nm was also demonstrated with an aperture size of 0.5×0.5 mm2.

But eventually, to realize not only an OPA but full LIDAR functionalities on a chip, active components like lasers, modulators and detectors need to be integrated together as well, and only modulators and detectors are available in silicon photonics. Hence, despite featuring higher optical propagation losses when compared to SiN, III-V materials like InP enable those active functionalities, and can be combined with SiN or silicon photonic PICs by using hybrid or heterogeneous integration.

As an example, a monolithic implementation of an OPA and tunable sampled-grating distributed Bragg reflector laser in InP achieved 12º in the lateral direction and 6º in the longitudinal direction [4]. Longitudinal beam steering controlled by the input wavelength demonstrated an efficiency of 0.14º /nm and very fast beam steering (>107 º/s) in both dimensions. Alternatively, an example of a hybrid implementation of a FMCW LIDAR PIC has been demonstrated using three distinct technologies are integrated into a single chip: active III-V (MEMS tunable VCSELs with high-index-contrast grating mirrors), Silicon Photonics (photodetectors, silicon photonic waveguides, couplers, and interferometers), and CMOS electronics (phase locked loop, temperature compensation, signal processing) [5].

Areas to explore further

While several photonic material platforms are mature enough to implement practical demonstrators of different integrated LIDAR systems, there are still challenges to improve their performance relative to their bulk counterparts. More effort needs to be focused on hybrid or heterogeneous integration, and on the scalability in design complexity, fabrication volumes and overall costs to be able to not only serve specialized niches, but also mass markets. A good evaluation of design trade-offs, foundry partners and packaging options is required when targeting a LIDAR integration project for a specific application.

Technobis Fibre Technologies started developing interrogators for FBG sensing in 2006. Following a successful development and launch of our Deminsys system, we started developing the most sensitive FBG interrogator in the world for ASML. This forced us to use integrated photonics to achieve stability in the sub-femtometer (10^-15 m). Using Multi-Project Wafer (MPW) runs in several  foundries, we developed the chips for ASML and at the same time our own Gator interrogator products.

Now, after 24 MPW runs and eight dedicated full-wafer runs, 64 designs, and more than 3,600 chips produced, launching serial production of the Gator, MultiGator, AeroGator, SwitchedGator, SwitchedRefGator, and Chiroptera, we can rightfully claim to be the first company in the world that is able to develop and supply sensor systems based on integrated photonics. This platform brings the most accurate interrogators, solid state, high speed, high sensor count, and low power systems, combined with unmatched system pricing.

This is recognized worldwide and brings us a large number of new customers in the areas we choose: Aerospace, Medical, High-tech systems, and Energy.

This also means that companies are diving into this new technology in an attempt to copy what we have, or to use the technology for other sensing applications.

We as Technobis are convinced that the market for sensing is too large for Technobis alone. At the same time, we need to ramp up the ecosystem in the Netherlands to enable the technology to mature to the desired level.

We believe that there is only one way to do that: by cooperating with this entire ecosystem and helping to grow the number of applications for sensing and data/telecom.

Therefore, Technobis started to offer the knowledge of designing and packaging PIC-based sensor systems to partners and (potential) competitors.

This will mean more designs, more design simulation, more MPW runs, more tests and analysis of the chips, more packaging, faster development of test equipment, tools, machines for assembly, etc.

Technobis strongly supports the ecosystem initiatives such as Photon Delta, the Institute for Photonic Integration, and the Photonic Integration Technology Centre.

Technobis will support this ecosystem with its testing, validation, and packaging foundry in Alkmaar.

• Generic & Custom Packaging
  
• Support for InP, SOI an TriPleX PICs and modules
  
• Semi- & Fully-automated packaging equipment
  
• Dedicated and extremely experienced team
  

The first two partnerships started over the past months with team-ups with companies that could have been competitors in the future: one on a different kind of FBG sensing for applications in a very specific field of applications, and the other company on interferometric sensor systems.

We are convinced that Technobis can help these companies reduce the number of chip design iterations significantly, evaluate the chips, and package the chips in such a way that the functions of the chip remain intact.

Technobis can shorten the development trajectory by years, helping us all to ramp up the PIC ecosystem in Holland in the coming years.

With an estimated 3-4 billion installed fluorescent tubes throughout the world, the integration of built-in LiFi transmission technology in new and retrofit LED light bars is now moving LiFi beyond the pilot stage to full-scale implementation in offices, schools, warehouses and other facilities.

“LiFi is not a concept, it is really here,” says Harald Haas, co-founder and chief science officer of pureLiFi, a company that is spearheading the development of the technology. “If people want to engage, they can purchase the products right now.” 

What is LiFi?

LiFi is a high-speed, secure, fully networked wireless communication technology similar to Wi-Fi. However, LiFi utilizes the entire light spectrum where Wi-Fi utilizes radio frequencies (RF). 

To do this, the LED light fixtures used in many energy-conscious homes and offices are outfitted with a module that controls the light for optical data transmission. The high-speed light pulses are invisible to the naked eye, yet can be used to transmit data rapidly to a receiving device located in a laptop, computer tower, cell phone or other smart device. 

Mobile opportunities: LiFi developers report that major mobile device manufacturers are expecting to adopt the technology within the next 3-5 years.

Lifi networks could also play a key role in machine-to-machine communication and the Internet of Things (IoT). 

The utilization of light provides a host of intriguing benefits. When compared to the overloaded full RF spectrum, the light spectrum is 1,000 times larger and is currently unregulated with no licensing fees. 

In lab conditions, the technology is already capable of 10 Gbps speeds, and with the available bandwidth potential, data transmission speeds up to 100 times faster will be possible in the near future as the technology advances.

Manufacturing push

According to Haas, who is considered the “father of LiFi” and has been working in the field for the past 15 years, the implementation of the technology into lighting fixtures has necessitated a close partnership with LED light manufacturers.

“The lighting manufacturers are very important to move LiFi forward,” says Haas. “They know how to design lights and fixtures and we know what needs to be done to create high-speed data networks out of light and add communication capability to it.”  

Until recently, most of these fixtures were small lamps or recessed can lights. Now, one of its partners, Linmore LED, is introducing the first LiFi enabled LED light bars designed to replace fluorescent tube lighting. For those that want to experience the technology in action, the company is demonstrating a complete, functional LiFi system using the new linear LED light bars at its facility in Fresno, Calif.  

With the technology, data speeds have been clocked at 43 Megabits-per-second (Mbps) up and down. 

“Linmore LED is the first company in the world to bring this technology not only into new light bar fixtures, but also be able to retrofit linear fluorescent fixtures that employ the LiFi technology,” says Haas.  

Linmore LED originally built its reputation in the retrofit market, utilizes its own proprietary designs involving optics, thermal dissipation and a number of other techniques to ensure its LED products perform in the top 1% in energy efficient in the industry. 

According to Paul Chamberlain, CEO of Linmore LED, the partnership with pureLiFi was a good fit due to the modular nature of the company’s LED light bars. The product’s design allowed for the integration of the LiFi modules in the ideal position on the light bar, without affecting critical aspects such as lighting distribution, thermal dissipation or overall performance.

Retrofit fixtures, even those that are not LiFi enabled, are in great demand as many facilities seek to drive down energy costs by as much as 70-80% by converting to LED technology. This trend is also being driven by the increased operating life of LEDs and concerns about the toxic mercury utilized within fluorescent lamps that complicates disposal.

This provides a very real scenario where building owners and facility managers can adopt LiFi technology while dramatically decreasing lighting-related energy costs at the same time.

“Businesses want to leverage an LED upgrade and get more than just lighting, says Paul Chamberlain, CEO of Linmore LED. “Utilizing an existing part of a building’s infrastructure – lighting – opens up endless possibilities for many other technologies to have a deployment backbone. Internet of Things (IoT), RFID, product and people movement systems, facility maintenance, and a host of other technologies are taken to the next level with LiFi available throughout a facility.”

Security benefits

Among the expected early adopters of the technology are those that seek greater security of data transmission than is possible with Wi-Fi.  For this reason, initial markets expected to adopt LiFi technology include federal government and defence, banking, financial institutions and hospitals.

With LiFi, devices must be directly within the cone of light to receive information. Visible light, including near-infrared wavelengths, cannot penetrate opaque objects such as walls, which means that the wireless signal is constrained to within a strictly defined area of illumination, which offers security benefits.

Wi-Fi, on the other hand, utilizes radio waves that are widely broadcast even outside a building where it can be easily intercepted for malicious purposes.

In a man-in-the-middle attack the attacker must be able to intercept all relevant messages passing between the two victims and inject new ones. This is straightforward in many circumstances; for example, an attacker within reception range of an unencrypted wireless access point (Wi-Fi) can insert themselves as a man-in-the-middle.

Because visible light is easily containable within a space, it could eliminate classic man-in-the-middle attacks where eavesdroppers located outside an area are able to intercept communications from radio waves emanating outside building.

In addition, traditional encryption and authentication protocols used for Wi-Fi provide an additional layer of security for the LiFi network.

The ability to direct or shape light into defined areas of illumination allows precise partitioning of any environment.  

File access is permitted only if a device is connected to the LiFi network. Once a user connects to the LiFi network, they can download and modify certain files. It is also impossible for a nearby employee to intercept information sent to the server/network by another employee, since the uplink communication is on a different frequency from the downlink.  

Further increasing security, every device that can connect to the network can be localized and tracked using the technology. The same LiFi module enables “communication on the move” by tracking the transmission source electronically, with no moving parts.

“You can walk through a building, into different [light] zones and it will keep you connected the entire time as you move along in the building,” says Haas.

The future

Now that one of the final barriers to full-scale implementation has been overcome with the introduction of LiFi enabled LED light bars, the technology is expected to continue to advance under an “aggressive strategy of miniaturization and lower costs,” says Haas.   

Delegates at the recent European-Japan PIC meeting, which took place in Tokyo (7-10 November), met to discuss a range of topics key to advancing photonic integration. As a main application of PICs in mass-markets, datacom-telecom was a major focus of the workshop. Standardization was another important theme regarding the mass-volume manufacturing of PICs and the realization of cost reduction opportunities. The PIXAPP Pilot Line for packaging and assembly of photonic devices was appointed as a key platform for the development of standardized packages. Let’s take a look at some of the technologies presented together with further highlights and conclusions from the meeting. 

Attendees had the opportunity to discuss a wide range of topics including the evolution of silicon photonics.

Telecom/Datacom

The first presentation was given by PETRA; the focus of PETRA (Photonics Electronics Technology Research Association) is the research and the development of Photonics and Electronics Convergence Technologies. The group updated the audience on its Integrated Photonics-Electronics Convergence System Technology Project (PECST-Pj) with the goal of reducing the size of current server rack equipment. The vision here features “on-board” datacentres with reduced power consumption. 

The PhoxLab team showed its approach to enhancing photonics in datacentres: PhoxTrot, which offers a product portfolio for converged storage, compute and ultra-high radix switching systems in object oriented fully optically disaggregated data centres. PHOENIX+ presented their activities related to the initiation of primarily R&D-related, but also business-related contacts between companies and research institutions in Berlin and partner regions. 

The BiCMOS Technology for High-Speed Photonic-Electronic Integration presented by Ihp attracted much interest among Japanese companies. This technology, which is based on electronic-photonic integration in SiGe, offers higher switching speed, higher gain, low input imped, higher leakage and low integration density combined with the advantages of CMOS technology. 

During the workshop, chip-to-fibre coupling was identified as a target area for developers. OFS presented solutions for optical interconnects supporting distances over 500 m. In collaboration with Furukawa, they are combining fibres with connectors and device assembly, offering photonic integrated circuit connection. Also, the trends of packaging and interconnect technology of Fujitsu SPARC64 Processor, which is part of the mother board of the K-computer, and the improved performance when using 2.5D technology for photonic-electronic integration were discussed. Fujikura presented a silicon DP-IQ Modulator for digital coherent transmission operating by low driving voltage (<2 Vppd) without RF amplifiers.

PIC design and simulation

PIC component development requires simulation and layout tools, not only to predict the optical performance of the device, but also the electrical and thermal behaviour. During its presentation, Lumerical provided information about photonic and optoelectronic TCAD device simulations and photonic integrated circuit design products. VPI Photonics showed the application range that its software can afford from waveguides and fibres to engineering and equipment configurations. This includes circuit simulation and verification considering critical layout constraints.

InP photonics

Fraunhofer HHI presented its InP Monolithic MZ modulator. Various examples of the HHI InP Foundry by Multi Project Wafers (MPW) were also shown. Smart Photonics introduced the audience to its PICs on InP technology based on monolithic integration of all photonic functionalities such as passive waveguides, phase modulators, gratings and amplifiers/laser gain section. Some applications of this technology were then demonstrated such as a widely tunable laser. InP PIC platforms of Smart Photonics and Fraunhofer HHI were considered of high relevance and a clear alternative to hybrid integration of silicon photonics.

Silicon photonics

ePIXfab – the European Alliance that promotes a fabless model for the development of silicon photonic circuits based on existing photonics and CMOS infrastructure – presented the different silicon-based fabrication platforms available in Europe and their respective capabilities. Looking at other projects, the goal of Cornerstone is to establish silicon photonics fabrication capability that can support photonics research in the UK via MPW. During their presentation, they showed their different technologies such as wafer-scale processing (DUV photolithography), chip-level processing (e-beam lithography) and ion implantation. The National Institute of Advanced Industrial Science and Technology (AIST) highlighted silicon-photonic integration activities at SCR-AIST and recent efforts toward industrial demands.

VTT presented its silicon photonic technology characterized by low losses and a small polarization dependency. These results are achieved by the combination of two complementary waveguide structures on 3-12 µm SOI, and feature III-V hybrid integration on SOI instead of monolithic integration. NTT showed the fabrication of heterogeneously integrated membrane III-V devices on Si.

Polymeric waveguides

A high thermally stable material for molding techniques, named SUNCONNECT was presented by Nissan Chemical. This polymer can fabricate optical waveguides in combination with the mosquito method (a dispenser). Kyocera Corporation presented technologies for optical and electrical wiring substrates including a polymer optical waveguide, turning mirror and optical card edge connector.

Sumitomo Bakelite presented multimode waveguide applications in high-speed data transmission. They are using a polymer waveguide for the PMT and SB Connectors.

PIC assembly and testing

The evolution of automated PIC assembly and testing as requisite for increased production volumes and lowering the cost-per-part was presented by Ficontec. In their talk, Yenista optics discussed different test techniques for PICs. They see three levels of testing, the wafer-level in which hundreds of PICs are on a wafer, the hybrid-level in which the PICs are mounted on electronics board and the packaged-level featuring fibre pigtailed opto-electronic components. 

Yelo presented a process of wafer qualification undertaken during device development known as accelerated ageing. It consists of testing a representative sample of devices under thermal and electrical conditions that accelerate the ageing process. Infant mortality failures are often caused by defects introduced during the manufacturing process. Accelerated aging can identify and remove such defective devices as early as possible in the manufacturing process. 

PI showed the reduction of the production alignment costs by 100x through parallelism. PI’s FMPA (fast multichannel photonics alignment) algorithm allows simultaneous alignment of all inputs and outputs and all DOFs. When you are aligning using parallelism to achieve typically 2 orders of magnitude speed improvement.

Final thoughts from the event

At the meeting, it was great to see photonic communities working together, interchanging ideas and knowledge. At EPIC, we can foresee a profitable Europe-Japan collaboration that could have a remarkable impact in the future of the PIC industry and PIC-based products.

Contact us on the details below to talk about your editorial, advertorial or advertising needs.

E: sukhi.bhadal@angelbc.com

W: www.picmagazine.net

T: +44 24 7671 8970