A new study has revealed the development of scalable photonic integrated circuits using the inorganic compound lithium tantalate (LiTaO3), a diamagnetic solid that is insoluble in water. Researchers at EPFL have made significant advancements in optical technologies with potential for widespread commercial applications.
A photonic integrated circuit (PIC) is a microchip that contains photonic components utilizing light or photons rather than electrical components like transistors, inductors, and resistors, which facilitate electron flux. Photonic chips employ optical components like lasers, waveguides, and polarizers to manipulate photons.
A photonic chip harnesses photons instead of electrons to process and distribute information. The use of light in place of electricity allows such technology to overcome electronics” limitations, such as heat generation.
This allows PICs to offer the advantages of low thermal effects, large integration capacity, miniaturization, and higher speed. At the same time, PICs are compatible with existing processing flows, allowing for lower prices, volume manufacturing, and high yield.
As a result, PICs found their application in a broad range of industries, from automotive to astronomy, sensing, biomedical, and data communications. As scientists and researchers continue to work on PICs to improve upon them, they can see even more applications and mark a new technological era.
To date, PICs have revolutionized computing systems and optical communications. However, silicon-based PICs have been leading the market so far due to their cost-effectiveness and ability to integrate with existing semiconductor manufacturing technologies.
Despite the electro-optical modulation bandwidth limitations of silicon-based PICs, silicon-on-insulator optical transceiver chips have been widely commercialized. However, recently, lithium niobate-on-insulator (LNOI) wafer platforms, known for their superior Pockels coefficient essential for high-speed optical modulation, have started gaining attention.
Electro-optical PICs based on lithium niobate (LiNbO3) exhibit vast capabilities due to their strong Pockels coefficient, finding use in high-performance computing, photonic accelerators for AI, and data-center communications.
However, there is a reason lithium niobate hasn’t seen wider adoption and commercial integration yet. The reasons are its complex production requirements, limited wafer size, and high cost per wafer, which are limiting its industrial use.
The technology is expensive because there are no existing high-volume applications like those that accelerated the adoption of silicon-on-insulator (SOI) photonics.
In the past two decades, silicon-based PICs have rapidly transitioned from academic research to widespread use in data centers, driven by the cost-effectiveness and high-volume availability of SOI wafers. Prepared using smart-cut techniques, these SOI wafers have enabled the manufacture of silicon photonics and, more importantly, are widely used in consumer microelectronics.
To put it in context, over 3 million SOI wafers with a diameter of 300 mm are produced globally every year.
However, thanks to the latest study, which used lithium tantalate (LiTaO3) for electro-optical PICs, things may finally change for them.
LiTaO3, similar in crystal structure to LiNbO3 but with heavier Ta atoms replacing Nb, offers greater mass density and stronger chemical bonds, enhancing strength and chemical stability. Its larger optical bandgap also enables nonlinear optical conversion to the visible and even ultraviolet wavelength range and shows a greatly decreased optical anisotropy, suppressing mode mixing.
Moreover, the modulation efficiency of both materials is expected to be almost identical, but LiTaO3’s larger optical damage threshold makes it very important for high-power applications.
Already widely used in commercial 5G radiofrequency filters, LiTaO3 is projected to reach a production capacity of 750,000 lithium tantalate-on-insulator (LTOI) wafers a year, potentially enabling the low-cost scalable manufacturing of LiTaO3-based PICs.
LiTaO3 PICs for Scalable Manufacturing
The latest study, published in Nature and funded by the Swiss National Science Foundation (SNSF) and the European Research Council (ERC), is making progress in achieving scalability in the production of large-volume and low-cost electro-optical PICs that are advanced and will cater to the future world in a better way.
This was made possible with the help of lithium tantalate, a material closely related to lithium niobate (LN) — a multifunctional ferroelectric crystal consisting of oxygen, lithium, and niobium which is the foundation of modern photonics and is a key material for optical modulators, mobile phones, optical waveguides, and piezoelectric sensors.
Like lithium niobate (LiNbO3), lithium tantalate (LiTaO3) has fantastic electro-optic qualities, but it also has additional benefits: cost and scalability. LiTaO3 further exhibits a much lower birefringence, enabling high-density circuits and broadband operation over all telecommunication bands compared to LiNbO3.
Given that the telecom industry is already using this material extensively in 5G radiofrequency filters, LiTaO3 can help overcome the barriers of lithium niobate.
So, the scientists at EPFL, led by Professor Tobias J. Kippenberg and Xin Ou, who’s a professor at the Shanghai Institute of Microsystem and Information Technology, created a PIC platform based on lithium tantalate.
This new PIC takes advantage of the LiTaO3’s inherent qualities to make high-quality PICs more economically viable, transforming the field of electro-optical PICs. The study demonstrated that lithium tantalate can be etched to create low-loss PICs by using a manufacturing process based on a deep ultraviolet (DUV) stepper as well as a lithium tantalate Mach–Zehnder modulator (MZM). The platform also supports the generation of soliton microcombs, which is a promising new approach for photonic-based microwave signal synthesis.
To achieve this great accomplishment, the team developed a wafer-bonding method for lithium tantalate, which works with SOI production lines.
The researchers fabricated LiNbO3 into lithium niobate-on-insulator (LNOI) structures to provide a completely new class of electro-optical PICs with extremely high speed and low voltage. According to the study, these PICs can become an integral part of future energy-efficient communication systems.
The fabrication was done using the smart-cut technique. This fabrication of LTOI is more closely aligned with the high-volume commercial production of SOI wafers, offering higher efficiency and lower production costs.
Next, the team masked the wafer with carbon and then began etching modulators, optical waveguides, and factor microresonators of supreme quality. They combined two techniques to etch the wafer: DUV photolithography, which involves creating patterns using controlled light, and dry etching, which involves removing material by exposing it to ions. Initially developed for lithium niobate, this etching process was later adapted to etch the harder and more inert lithium tantalate.
The etching was adapted to minimize optical losses, which is critical in achieving high performance in photonic circuits.
As a result, the researchers were successful in fabricating lithium tantalate PICs, which are highly efficient and have an optical loss rate of only 5.6 dB/m at telecom wavelength.
As noted above, electro-optic MZM — currently used in high-speed optical fiber communication — was another highlight of this study. The lithium tantalate MZM’s electro-optical bandwidth reached 40 GHz while offering a half-wave voltage-length product of 1.9 V cm. According to the study:
“Our process is fully wafer-scale and based on deep-ultraviolet photolithography and lays the foundation for the scalable manufacture of high-performance electro-optical PICs that can harness the scale of LTOI wafer fabrication for 5G filters.”
This way, the study’s LTOI PICs achieve similar loss and electro-optical performance to the well-established LNOI technology that has major potential for use in data-center interconnects, long-haul optical communications, and quantum photonics. According to Chengli Wang, the study’s author:
“The study was further able to generate soliton microcomb while maintaining highly efficient electro-optical performance. These soliton microcombs feature a large number of coherent frequencies and, when combined with electro-optic modulation capabilities, are particularly suitable for applications such as parallel coherent LiDAR and photonic computing.”
Extensive Real-world Applications of Lithium Tantalate PICs
The development of photonic integrated circuits (PICs) using lithium tantalate (LiTaO3) is a big achievement, given its numerous real-world applications across various fields. Some key areas that could benefit from this advancement include data centers, which can utilize these PCIs to manage massive amounts of data traffic and reduce latency.
In telecommunications, lithium tantalates are already being used. By integrating them into PICs, we can streamline the transition to cutting-edge wireless communication systems. Meanwhile, MZMs can improve data transmission rates in optical fiber networks.
Computing is another area where the ultra-high speed offered by these PICs can help the processing capabilities, while quantum computing can benefit from lithium tantalate PICs’ high efficiency and low birefringence. Consumer electronics is yet another field that can utilize lithium tantalate PICs for better device performance.
Then there is LiDAR (Light Detection and Ranging), which can be made more accurate for use in autonomous vehicles, agriculture, weather, navigation, biodiversity, mapping, and much more.
Other industries that stand to benefit include sensors, which are utilized in medical, industrial, and environmental monitoring, and healthcare, where advanced PICs can help with early diagnosis and treatment planning.
As we saw, leveraging lithium tantalate’s unique properties allows PICs to offer enhanced performance, scalability, and cost-effectiveness. These benefits make PICs based on lithium tantalate highly suitable for a wide range of applications that demand high-speed, high-efficiency optical solutions.
Click here to learn why lasers are set to play a pivotal role in the coming decades.
Companies That Can Benefit From This Advancement
The latest advancement in PICs has profound use cases, as we talked about above, now, let’s take a look at a couple of companies that can benefit from this new technology:
#1. Lumentum Holdings Inc. (LITE)
Lumentum is a provider of optical and photonic products, covering the lasers and communications market via the Lasers and OpComms segments. The company has a market cap of $3.12 bln, while its shares (LITE: NASDAQ) trade at $46.28, down 11.71% YTD. Its EPS (TTM) is -5.26, and its P/E (TTM) is -8.81.
For its recent quarterly results, Lumentum reported net revenue of $366.5 million, a bit higher than estimated. However, there was a decline from the previous quarter, in part due to inventory issues faced by telecom customers.
Non-GAAP net income was $19.6 million, while GAAP net loss was $127 million. GAAP gross margin meanwhile fell to 16.2%, and non-GAAP gross margin reduced to 32.6%. The company’s total cash, cash equivalents, and short-term investments also dropped to $870.9 million, mainly because of a $323mln repayment in convertible notes.
CEO Alan Lowe, however, is optimistic about the telecom sector’s recovery and has expressed confidence in the company’s ability to capitalize on the growing demand for AI-driven cloud data centers.
#2. Infinera Corporation (INFN)
Infinera Corporation supplies advanced optical semiconductors and optical networking solutions to enterprises, governments, carriers, and cloud operators. The company has a market cap of $1.27 bln, and its shares trade at $5.46, up 14.95% YTD. Its EPS (TTM) is -0.34, and its P/E (TTM) is -16.04.
For Q2 of 2024, the company reported $306.9 million in GAAP revenue, down from $453.5 million in the previous quarter. GAAP gross margin, meanwhile, was 36.0%, again a decrease from 38.6% in 4Q23 and 37.5% in 1Q23. For the quarter, the GAAP operating margin was 14%, and the GAAP net loss was $61.4 million.
The non-GAAP gross margin was 36.6%, the non-GAAP operating margin was 8.4%, and the non-GAAP net loss was $38.3 million. The company generated $24 million of operating cash flow and free cash flow of $16 million to end the quarter with cash, cash equivalents, and restricted cash at $192.2 million.
CEO David Heard called 1Q24 an important quarter, marked by significant customer, RFP, and design-win momentum.
Concluding Thoughts
While PICs based on silicon photonics have been leading the world with widespread usage in telecommunications and data centers in the past many decades, things are going to get even more interesting with the next generation of PICs making their entrance.
Ultrahigh-speed PICs based on electro-optical materials will be playing a bigger role in energy-efficient data centers, 5G and 6G radiofrequency filters, optical communications, and, most importantly, in AI workload-driven high-performance computing.
Of course, for that to happen, we need scalable, low-cost manufacturing, which the latest study has accomplished by using LiTaO3, which is, in some cases, superior to LiNbO3. Being compatible with existing production lines makes this new method extremely enticing. It promises low-cost next-generation electro-optical PICs with wide applicability across industries.
Click here to learn how new phononics research might revolutionize communication devices.