Hamamatsu Photonics has developed a scientific camera that is attracting the attention of researchers and engineers working on next-generation quantum computing architectures
There are incremental innovations (think continuous improvement of existing product lines) and then there are disruptive innovations (think platform technologies that rewrite the rulebook). A case study in the latter is the ORCA®-Quest, a scientific camera that has been turning heads since its commercial release in 2021, opening up cutting-edge imaging applications in disciplines as diverse as quantum computing, atomic physics, synchrotron science, Raman spectroscopy and super-resolution microscopy.
Developed by Hamamatsu Photonics, a specialist manufacturer of high-sensitivity, low-noise cameras for fundamental and applied research, the ORCA®-Quest is a quantitative CMOS (qCMOS) camera with unique “photon-number resolving” functionality – determining the number of photons incident on each pixel (in a 9.4 megapixel array) by accurately measuring the number of photoelectrons generated on a per-pixel basis.
Innovative design, advanced fabrication
It’s this granular ability to count photoelectrons that underpins the ORCA-Quest’s game-changing performance as an imaging system. “To realize photon-number resolving, we had to make some modifications to the pixel structure itself,” says Brad Coyle, OEM camera product manager at Hamamatsu Photonics.
Specifically, that means fabricating deep-trench structures in the semiconductor layers between each pixel to ensure that a photon that impacts on a given pixel registers exclusively on that pixel. “The trench structure suppresses the flow of photoelectrons between neighbouring pixels,” Coyle adds, “so we get really high fidelity and linearity at the level of the individual pixels.”
Equally important in this regard is the work of the Hamamatsu development team to reduce the noise-floor of the ORCA-Quest. Conventional scientific CMOS (sCMOS) cameras, for example, come with a low readout noise – though still larger than the photoelectron signal, which makes it difficult to count photoelectrons.
“What we found, theoretically, is that in order to quantify the number of photoelectrons generated per pixel, we have to reduce the noise floor below 0.3 electrons,” notes Coyle. In terms of practical implementation, this necessitated a redesign of the detection and readout circuitry of the ORCA-Quest to achieve 0.27 electron RMS read noise (while ensuring stable performance versus temperature and time as well as individual calibration and real-time correction of each pixel value).
Another notable feature of the ORCA-Quest is its high-speed readout – in other words, how many pixels the camera can read out per second (number of pixels × frame rate). In standard scan mode, for example, the ORCA-Quest offers a higher data rate (>1100 megapixel/s) and lower readout noise compared with conventional sCMOS cameras (approx. 400 megapixel/s). In ultraquiet scan mode, meanwhile, the camera offers photon-number resolving with a x10 faster data rate (approx. 250 megapixel/s) versus single-photon counting with electron-multiplying CCD cameras
“With the full 9.4 megapixel array [4096 x 2304] and high-speed readout, users are able to image a large number of objects with exacting temporal requirements,” notes Coyle. “The high-bandwidth interface means it is also possible to extract real-time feedback from the camera system – which is mandatory for emerging R&D applications in quantum computing and quantum communications.”
Quantum imaging, quantum insights
Among the early-adopters of the ORCA-Quest in the quantum science community is Dolev Bluvstein, a PhD student and team member of Mikhail Lukin’s Quantum Optics Laboratory at Harvard University (Cambridge, MA). Within a diverse programme of theoretical and applied research, Bluvstein and colleagues are working on aspects of quantum computing and quantum simulation using arrays of individually trapped rubidium-87 (Rb) atoms.
At a schematic level, individual atoms are trapped independently in vacuum by optical tweezers, such that highly focused laser beams enable real-time control of each atomic position in space. “Once the atoms are prepared in their programmed positions and pumped into their ground electronic states,” says Bluvstein, “we introduce interactions among them by using lasers to excite them to their Rydberg states [in which an electron is excited into a very large orbital state].”
In this way, the Rb atoms are aligned one-by-one in an array to be utilized as qubits for quantum computing operations, while the qubit states are determined by observing the laser-induced fluorescence (or absence of fluorescence) from each atom. It’s here that the ORCA-Quest provides a core building block in Bluvstein’s experimental set-up, ensuring spatial diagnostics of the entire atom array as well as quantum-state detection for each atomic qubit – all while combining ultralow-noise measurements and high-speed readout (at a frame rate of every 100 μs).
“The camera is the only way to see where the Rb atoms are and to extract qubit information rapidly out of the quantum system,” says Bluvstein. “In a sense, the camera is the main input/output interface between the classical and quantum worlds.”
At the end of last year, Bluvstein and colleagues published a landmark paper in Nature, detailing work on a quantum processor architecture based on reconfigurable atom arrays. The laboratory testbed – which was developed as part of Bluvstein’s PhD work within a wider collaboration involving scientists at NIST/University of Maryland, Massachusetts Institute of Technology (Cambridge, MA) and QuEra Computing (Cambridge, MA) – features high-fidelity entangling gates, local qubit control, mid-circuit readout and any-to-any connectivity for hundreds of atomic qubits.
Silicon photomultipliers: gearing up for applications in gamma-ray astronomy
By grouping atomic qubits together to form error-corrected logical qubits, the team is exploring early fault-tolerant quantum computation with up to dozens of logical qubits and hundreds of logical entangling gates. The ultimate end-game: a neutral-atom, error-corrected quantum computer – at scale – with of the order of 10 million atom qubits imaged every 100 μs on a 24/7 basis while logging their individual quantum states.
“The ultrafast readout speed of the ORCA-Quest over large regions of interest is mission-critical for our research,” explains Bluvstein. “What’s also impressive is that when we installed the camera in our experimental set-up, the signal-to-noise of our atomic imaging improved by a factor of two. All of which advances the rate at which we can do this classical/quantum interfacing and, ultimately, the path to large-scale quantum computation.”
Continuous improvement
Disruptive innovation, inevitably, begets incremental innovation. Earlier this year, Hamamatsu Photonics unveiled the ORCA-Quest 2, offering end-users enhanced functionality along a couple of key coordinates. For starters, there’s a x5 improvement in frame rate when the camera is operated in ultra-quiet scan mode for photon-number resolving (i.e. 25 frame/s for the full 4096 x 2304 array). The ORCA-Quest 2 also offers higher quantum efficiency in the UV region (around 280–400 nm) thanks to an advanced antireflection coating on the sensor window (with no change to the efficiency in the visible and near-IR regions).
It’s early days, however, and the search is on for new applications and market segments for the ORCA-Quest 2. “We’re looking to hear from partners who may have a unique application for this camera,” concludes Coyle. “This is still a new technology and we’re trying to learn where the best fit is going to be across basic research and industry R&D.”
Further reading
D Bluvstein et al. 2023 Logical quantum processor based on reconfigurable atom arrays Nature 626 58
