Recently, the Dutch Research Council (NWO) revealed six research projects to receive funding within its Open Technology Programme. We are proud to announce that our project – Atomic Layer Deposition and Etching for Quantum Computing – is among these projects! In this blog we will outline why this project was conceived, and why the addition of atomic-scale processing (ASP) to quantum technology (QT) nanoscale fabrication techniques is such an exciting and promising development.
Quantum technology: Why is it so revolutionary?
There is a strong worldwide interest in quantum technology (QT) for computing, communication, and sensing, driven by the potential leap in capabilities made possible by controlling the quantum states of matter. Previously unattainable feats such as the simulation of extremely complex systems in chemistry and materials science and the development of detailed weather models for climate investigation are made possible by the introduction of QT. The most well-known realization of QT is the quantum computer. Whereas classical computers use bits, which are either 0 or 1, to execute computations, quantum computers use qubits. These qubits are two-level quantum systems, meaning they can be 0, 1, or any superposition of the two states. This allows for exponentially faster computations compared to their classical counterparts. Most popularly, the quantum mechanical behavior of qubits is achieved through superconductivity. However, the quantum mechanical state of a qubit can be lost due to interactions with environmental degrees of freedom: noise. This phenomenon is known as decoherence, and the lifetime of the qubit’s quantum mechanical behavior is given by the coherence time. The energy loss to the environment can be expressed in terms of the qubit’s quality factor Q ~ 1/loss. Important sources of loss which can be affected by choice of materials and fabrication methods are depicted in the figure below for the example of a superconducting flux qubit. Examples of noise reduction through materials and processing are minimizing the impurity and defect density, reduction of lossy oxides, and improving layer homogeneity and interface quality.
Materials science to bridge the gap to large-scale quantum systems
Decoherence is unavoidable, as the qubit cannot be completely isolated from the environment. This limits the number of gate operations which can successfully be performed in one quantum algorithm. To optimize device performance, loss to the environment must be reduced as much as possible. There are two ways to do so: reducing the qubit’s noise sensitivity through design choice and reducing the noise itself through material and fabrication optimization. While there has been much progress in qubit design optimization to circumvent noise sources, sources of loss have scarcely been addressed beyond material selection. Hence, there resides a great deal of unused potential in the materials science community for the improvement of quantum device quality and scalability. As stated by Oliver & Welander (2013) :
Realization of large-scale quantum systems, for which error rates lower than currently demonstrated must be achieved, requires major material advances . De Leon et al. (2021)  emphasize that many problems faced by the QT community resemble those addressed by the semiconductor industry over the past five decades:
We envision tailored ASP approaches as enablers of large-scale QT device fabrication, like they were in recent generations of nanoelectronic devices, as illustrated in the figure below.
An example of the new possibilities opened up by using the tools of materials science for quantum devices is shown in the figure below [1, 3]. The slide depicts a trend similar to Moore’s law for the number of transistors in an integrated circuit. In this case, qubit coherence times leaped from nanoseconds to nearly milliseconds between 2000 and 2020. Currently, the industry is closing in on the millisecond benchmark, with coherence times in the tenths of milliseconds having been achieved for tantalum 2D transmon qubits (another type of superconducting qubits) in 2021 and 2022 [4,8]. Different qubit designs, which are shown to be the dominant instigators of coherence time improvement, are indicated in the figure on the left of the slide. This overview by Oliver et al., albeit not very recent, remains very relevant concerning the key role of materials research in furthering superconducting qubits to new frontiers. In the 2010s the introduction of 3D qubits led to record coherence times. However, 3D qubits are not widely used due to their large size. Strikingly, for the more favorable 2D qubits a large jump in coherence times rivalling those of 3D qubits was achieved through material advances. This begs the question: what further jump in coherence times through materials and fabrication improvements still await discovery?
The current record qubit coherence times are still well below the theoretical maximum limit predicted by bulk dielectric loss . Surfaces and interfaces thus seem to play a significant role as sources of loss. This, together with the coherence time improvement achieved through material changes, e.g. by use of tantalum , demonstrates that material and interface optimization/passivation are key to enhancing qubit performance.
Though there is no question that materials development and nanoscale fabrication will play a key role in future quantum device production, so far little work has been done to reach their full potentials. With the rapidly increasing technological readiness of QT – illustrated by the large investments of many companies and governments – the field has to start focusing on scalable fabrication methods for qubits and other quantum devices. Our project aims to contribute to this development by taking our ALD/ALE toolbox aimed at the semiconductor industry and tailoring it to the specific needs of various quantum applications. With its unique combination of capabilities, we anticipate ASP to become a key instrument in QT fabrication.
Why ASP for QT?
QT imposes strict requirements on materials and fabrication techniques. A very small amount of noise can already lead to decoherence. Operation at the cryogenic temperatures and single photon limit generally required for QT affects the relative importance of loss mechanisms, which introduces unique demands very different from those in the semiconductor industry. The ASP techniques of atomic-layer deposition (ALD) and atomic-layer etching (ALE) are very promising in this context, not only for their true atomic scale control, but also for compatibility with wafer-scale fabrication and high-volume manufacturing.
In recent years, ALD of superconducting refractory metal-nitrides (mostly NbN) has been explored for quantum applications because of the increasingly demanding requirement of precise (ultrathin film) thickness control combined with high uniformity in quantum devices . The number of studies on this subject, however, remains comparably small. In the field of quantum sensing, superconducting detectors are of high interest due to their wide range of applications from photon and particle detection to astronomy . To reliably detect photons and conform to the ever smaller on-chip features, the aim is to prepare ultrathin superconducting layers while maintaining layer quality.
Recently, Sheagren et al. (2020)  investigated the use of ALD NbN for nanofabrication of superconducting detectors. They were able to obtain film thicknesses much smaller (<10 nm) compared to sputtered films while maintaining comparably high quality factor (Q) values. TiN is also a material of high interest due to its higher photon sensitivity at large wavelengths (~µm). Shearrow et al. (2018)  fabricated microwave resonators from ALD TiN films between 9 and 110 nm with enhanced single photon coupling strengths for quantum sensing. In addition to metal nitrides, metal oxides are widely used in quantum devices. An example is the dielectric layer in the JJ, as shown in the first figure in this blog. Current fabrication methods such as double angle evaporation suffer from local thickness variations and fabrication decoherence sources such as photoresist residues and processing damage. Double angle evaporated JJs are also sensitive to aging affects, and are difficult to precisely reproduce, both of which negatively impact device yield. This is especially the case for devices with large numbers of JJs, such as those with high qubit counts. ALD of Al2O3 has shown promising results for the fabrication of ultrathin low-defect tunnel barriers in JJs , but more systematic research is needed to establish ALD as a key element for improving JJ performance.
Initial experiments in QT, borrowing existing ALD and ALE approaches from nanoelectronics, have alluded to the promise of ASP. To reach its full potential, new approaches must be developed, requiring a strong background in the field of ASP, to meet the unique demands of QT imposed by operating conditions (often cryogenic) and principles. The development of these new approaches will require interdisciplinary collaboration between the fields of materials science and quantum computing.
As material science ventures for QT are still in their infancy, many open questions are waiting to be answered. Which material combinations yield the best quantum performance with respect to the various critical interfaces? What precisely is the right combination of material properties to strive for, and how can the various ALD/ALE components be optimized to achieve these properties? What substrate preparation and device processing methods are best for QT applications? We aim to address these questions and many more together with our project partners. The latter are listed in the project slide below. Finally, before kicking off our project, we can already bring very exciting news! From exploratory research on superconducting metal nitride nanolayers, we report the highest critical temperatures (Tc) for tantalum nitride prepared by plasma ALD. The preliminary findings from this research have been presented at the ALD/ALE 2022 conference in Ghent (see slide below). Given our ambition to establish ALD and ALE as key fabrication technologies in the quantum revolution, this is a good start!
 Oliver, W., & Welander, P. (2013). Materials in superconducting quantum bits. MRS Bulletin, 38(10), 816-825. doi:10.1557/mrs.2013.229
 De Leon, N. P., Itoh, K. M., Kim, D., Mehta, K. K., Northup, T. E., Paik, H., Palmer, B.S., Samarth, N., Sangtawesin, S., & Steuerman, D. W. (2021). Materials challenges and opportunities for quantum computing hardware. Science, 372(6539), eabb2823. doi:10.1126/science.abb2823
 Gil, D., & Green, W. M. J. (2020). 1.4 The Future of Computing: Bits + Neurons + Qubits. 2020 IEEE International Solid- State Circuits Conference – (ISSCC), 30-39. doi: 10.1109/ISSCC19947.2020.9062918.
 Place, A.P.M., Rodgers, L.V.H., Mundada, P. et al. (2021). New material platform for superconducting transmon qubits with coherence times exceeding 0.3 milliseconds. Nat Commun ,12, 1779. doi:10.1038/s41467-021-22030-5
 Sheagren, C., Barry, P., Shirokoff, E. et al. (2020). Atomic Layer Deposition Niobium Nitride Films for High-Q Resonators. J Low Temp Phys 199, 875–882. doi:10.1007/s10909-020-02336-2
 Shearrow, A., Koolstra, G., Whiteley, S. J., Earnest, N., Barry, P. S., Heremans, F. J., … & Schuster, D. I. (2018). Atomic layer deposition of titanium nitride for quantum circuits. Applied Physics Letters, 113(21), 212601. doi:10.1063/1.5053461
 R. Lu et al. (2013). Fabrication of Nb/Al2O3/Nb Josephson Junctions Using In Situ Magnetron Sputtering and Atomic Layer Deposition. IEEE Transactions on Applied Superconductivity, 23(3), 1100705. doi:10.1109/TASC.2012.2236591.
 Wang, C., Li, X., Xu, H. et al. (2022). Towards practical quantum computers: transmon qubit with a lifetime approaching 0.5 milliseconds. npj Quantum Inf, 8, 3. doi:10.1038/s41534-021-00510-2