Periodic Reporting for period 3 - COINFLIP (Coupled Organic Inorganic Nanostructures for Fast, Light-Induced Data Processing)
Période du rapport: 2022-02-01 au 2023-07-31
COINFLIP seeks to make faster optical transceivers, which are crucial for every modern data-communication network, where they convert in-coming optical data from an optical fiber into electrical information to be processed by a computer end station. To excel in the speed of this transformation,this project exploits a team effort between inorganic quantum dots and organic semiconductor molecules which, when brought together, become responsive to very fast optical signals and show novel physical properties. For instance, we can greatly influence the way the material responds to a light pulse of one wavelength by manipulating it with a second light pulse of another wavelength.
A problem we are going to face in the near future is that data communication with optical fibers gets increasingly more efficient and faster, but the development of equally fast optical transceivers is lagging behind. This is somewhat like modernizing and expanding a countries’ network of motorways but neglecting the need for frequent and efficient exits. COINFLIP develops new materials for quicker optical transceivers to meet the future needs of a communication network, where a speed above 100 GHz is the expected norm rather than the exception. In this context, common materials for processing optical data input must find a compromise between speed and sensitivity. In simple words, a material cannot be fast and sensitive to an incoming signal simultaneously because registering a very weak stimulus requires some time to process it. COINFLIP merges one material, which is only fast, with a second material that is very sensitive (and slow). The sensitive component could be a dye, something that interacts with light so strongly that one could easily spot the smallest amount of it on a white wall. The fast material is a bit like a black hole to light: as soon as a light ray hits it, the light is gone. COINFLIP develops solution to make these two components work together effectively in order to result in very powerful optical transceivers. A major challenge towards this end is the measurement of electric signals that last only a few trillionth of a second (“picoseconds”). Therefore, a first milestone is to build a specialized infrastructure that is capable of tracking electric currents at this ultra-high speed. Once accomplished, this will be used to focus on the most promising materials and further improve their performance.
A problem we are going to face in the near future is that data communication with optical fibers gets increasingly more efficient and faster, but the development of equally fast optical transceivers is lagging behind. This is somewhat like modernizing and expanding a countries’ network of motorways but neglecting the need for frequent and efficient exits. COINFLIP develops new materials for quicker optical transceivers to meet the future needs of a communication network, where a speed above 100 GHz is the expected norm rather than the exception. In this context, common materials for processing optical data input must find a compromise between speed and sensitivity. In simple words, a material cannot be fast and sensitive to an incoming signal simultaneously because registering a very weak stimulus requires some time to process it. COINFLIP merges one material, which is only fast, with a second material that is very sensitive (and slow). The sensitive component could be a dye, something that interacts with light so strongly that one could easily spot the smallest amount of it on a white wall. The fast material is a bit like a black hole to light: as soon as a light ray hits it, the light is gone. COINFLIP develops solution to make these two components work together effectively in order to result in very powerful optical transceivers. A major challenge towards this end is the measurement of electric signals that last only a few trillionth of a second (“picoseconds”). Therefore, a first milestone is to build a specialized infrastructure that is capable of tracking electric currents at this ultra-high speed. Once accomplished, this will be used to focus on the most promising materials and further improve their performance.
We have tested a variety of different nanomaterials and explored their potential towards application in fast optical transceivers. This includes PbS, CuS and CdSe nanocrystals as well as Au nanoclusters. Most notably, we found that the surfaces of CdSe (https://arxiv.org/abs/1904.04752) and CuS nanocrystals (https://arxiv.org/abs/1903.05037) can be decorated with organic dyes to greatly expand their sensitivity for optical wavelengths which are typically not detected by these materials. In contrast, three-dimensional supercrystals of Au nanoclusters were found to exhibit highly interesting electric transport properties owing to the exceptionally large degree of long-range structural order (https://doi.org/10.1038/s41467-020-19461-x) but they turned out to be chemically too unstable under intense light pulses. However, this discovery has enabled a follow-up study which utilizes this material for the cost-effective fabrication of electrical micro-circuits. Following a similar idea, we have established a new method to print thin stripes of nanocrystal films onto electrical micro-circuits for further use as optical transceivers (https://arxiv.org/abs/2006.11202). To complement the findings on optical transceivers based on nanocrystal/organic dye hybrid materials, we have detailed the interaction between these two material classes in experiment (https://arxiv.org/abs/2005.00898) as well as with an extensive review of the field (https://doi.org/10.1002/anie.201916402). In our most important work for this project so far, we have designed optical transceivers based on CdSe nanocrystals and organic dyes and investigated the effect of different substrates onto the speed of their performance (http://arxiv.org/abs/2109.02049). We realized a 3dB bandwidth of 85 KHz and identified the relatively large dimensions of the lateral optical transceiver as the main cause for the current speed limitation. Further miniaturization should lift these limitations and lead to faster device performance. To this end, we have successfully put into operation a measurement station which can reliably measure the speed of an optical transceiver with a resolution of 1 ps. This new piece of equipment enables us from now on to measure bandwidths of up to 1000 GHz, which is sufficient to monitor the successful implementation of the key objective of this project, which is a response time of < 5 ps (>200 GHz). First results on PbS-based optical transceivers have resulted in an intrinsic response time of < 500 ps, which we view as an encouraging starting point to the second half of the project.
Compared to commercially available optical transceivers on the basis of macroscopic silicon or galliumarsenide or even the recently explored graphene, the speed of operation achieved in this project so far cannot yet surpass the state-of-the-art. However, the versatility of the approach by combining nanocrystals with dyes - which is not easily adaptable to the aforementioned materials - holds room for many improvements. Since the speed of lateral optical transceivers on the basis of nanocrystals is rather seldomly explored (in contrast to maximizing the sensitivity), an important step beyond the state of the art has been the identification of a suitable substrate for the devices. We found that many previous studies have reportedly utilized substrates (silicon covered with approx. 230 nm oxide) with questionable suitability for determining the speed of an optical transceiver due to a possible contribution of the (unwanted and unnoticed) signal of the silicon underneath the nanocrystal layer. Another important step beyond the state of the art has been the successful installation of a piece of equipment ("asynchronous optical sampling") which enables precise response measurements up to THz 3dB-bandwidths. Such measurements need to be carried out with an all-optical device since the electric leads to and from the sample to connect with an electric measurement station will falsify the measured response speeds. In the past, such measurements have been realised by pump-probe spectroscopy and so-called "Austin switches". This technique requires complex optical lithography and measurement times of several hours. In contrast, the asynchronous optical sampling technique now ermployed by us can be carried out on simple electric circuits and is completed within minutes. With respect to our finding that our current devices suffer from their relatively large size (2500 nm), we have started to use helium ion beam lithography to fabricate devices as small as 10 nm, which is amongst the smallest devices that have ever been tested.
We are confident that the results of the first half of the project will allow us to quickly improve the speed of optical transceivers based on nanocrystal/organic dye hybrid materials towards the state of the art and target the key objective of the project of a response time of < 5 ps at the end of the second half. The greatly reduced device dimensions and the newly installed ultrafast measurement techniques will play a lead role in this respect.
We are confident that the results of the first half of the project will allow us to quickly improve the speed of optical transceivers based on nanocrystal/organic dye hybrid materials towards the state of the art and target the key objective of the project of a response time of < 5 ps at the end of the second half. The greatly reduced device dimensions and the newly installed ultrafast measurement techniques will play a lead role in this respect.