I thank Joan Rosell-Llompart for his historical testimony. I should merely add that his recollections are far more valuable than mine. First, because my memory leaves much to be desired. Also, and more importantly, because Joan was completely in charge of our group's program to make and test a shorter version of the Reischl DMA. I should also note that this program was started at Joan's initiative, when I reported that the tests of a shortened Reischl DMA at Duisburg by Andreas Schmidt-Ott and myself had not increased the resolving power as theoretically expected.

I am grateful to Dr. Gerhard Steiner for his remarks. His endorsement of the program I was proposing is especially welcome given his position of responsibility for instrument development at Grimm.

It is most unfortunate that Georg Reischl, like Stradivarius, passed away without elaborating on the design criteria used in his DMAs. Gerhard's testimony relating to Georg's views on the importance of accelerating the sheath gas past the laminarization screens is therefore a valuable direct confirmation of a principle observable in the genuine Reischl prototypes. I have nevertheless a slight disagreement with Gerhard's account. In my own recollection, my only conversation ever with Georg was not by telephone, but in person, during a European aerosol conference, probably at Karlsruhe. There were a number of other participants in that (for me) memorable discussion, and I suppose Gerhard was one of them.

Dr. Steiner brings up the very important subject of the potential of using electrical detectors as alternatives to CPCs. This view is stated in general, not just in relation to high flow DMAs, though, in our application, electrometers present the obvious advantage of readily accommodating large flow rate variations.

Many readers may be skeptical about electrical detectors in atmospheric measurements, given that the typical electrometers used in aerosol research have noise levels of about 1-10 fA and response times of 1 s or more. These detectors thus require over 10^3 elementary charges/s (or particles/s). Nevertheless, there is ample room for improving the response time and the sensitivity of most commercial aerosol electrometers. We have reported the incorporation into a collecting filter of an operational amplifier circuit developed by Heinz Burtscher and his colleagues, which achieved a response time below 100 ms and a noise level of about 0.1 fA (Fernandez de la Mora et al. 2017). This electrometer was commercialized by the now bankrupt company SEADM, but a comparable variant is still obtainable through my former student Luis Perez-Lorenzo. Nevertheless, this improvement is still modest compared to what could be achieved in practice. For instance, CCD cameras typically convert individual photons into elementary charges, and then measure their current. Such rapidly evolving sensors could therefore be applied to electrical aerosol detection. That the single particle sensitivity of CPCs is in principle extendible to electrical measurements has been clear for some time. For instance, Ma et. al. (2017) state that "…  the 1.1 μm pixel-pitch device achieves 0.21e− rms average read noise with average dark count rate per pixel less than 0.2e−/s, and 1040 fps readout rate".  Note that the outstanding sensitivity of their CMOS-based photon-counting image sensor applies not just to a single detector, but to arrays of many detectors ordered in two dimensions.  This means that DMAs with such detector arrays would not need to scan over the voltage to obtain a size spectrum. The whole spectrum would be recorded almost instantly in a multitude of detectors distributed along the inner electrode. This possibility has already been demonstrated (Perez-Lorenzo et al. 2020) in a planar DMA with 100 operational amplifiers, though not with single charge sensitivity. This work shows that the carefully designed insulating steps separating the various metallic collectors do not cause turbulent transition. A commercial linear detector array based on CCD technology with a noise level of 0.5 fA has existed for some time (https://www.jas-sg.com/ids-2030-charged-particle-detector.html). It is 51 mm long, has 2126 active pixels each 21 µm wide and 1500 µm high. It has been used in mass spectrometry based on instruments that disperse the ions in space. As far as we know, it has not been tested with DMAs, so, whether or not the collector roughness would cause turbulent transition is unknown.

An array of electrical detectors with single charge sensitivity would be clearly superior to a DMA connected to a single high flow CPC, by a large factor equal to the number of mobility channels used. Note that the exceptional  low noise achieved by Fossum and colleagues relies on the small area of each pixel. A typical situation with 1 million pixels, each  1.1 micron in length would appear at first sight to divide the signal into too many mobility channels, each receiving too little current to be measurable in typical aerosol applications. Nevertheless, in view of the claimed ability by Ma et al (2017) of achieving single photon counting in each pixel, the signal would not be diluted over a wide mobility range, but assigned to precise values of the mobility. High mobility resolution would not imply in this case reduced sensitivity. One could certainly add the counts from as many pixels as desired by post-processing, without loss of the available high resolution information. Of course, the fact that such developments are in principle possible does not necessarily mean that they will be available soon.

J. Fernandez de la Mora, L.J. Perez-Lorenzo, G. Arranz, M. Amo-Gonzalez, H. Burtscher, Fast high-resolution nanoDMA measurements with a 25 ms response time electrometer, Aerosol Science and Tech., 51(6), 724 – 734, 2017

Ma, S. Masoodian, D. A. Starkey, E. R. Fossum, Photon-number-resolving megapixel image sensor at room temperature without avalanche gain, Optica,4 (12) 1474-1481 (2017) https://doi.org/10.1364/OPTICA.4.001474

Perez Lorenzo, R. O’Mahony, M. Amo-Gonzalez, J. Fernandez de la Mora, Instant Acquisition of High Resolution Mobility Spectra in a Differential Mobility Analyzer with 100 Independent Ion Collectors: Instrument calibration, Aerosol Sci. & Techn 54(10) 1144-1156, 2020

I thank the author of  RC1: 'Comment on ar-2023-7', Anonymous Referee #1, 29 Oct 2023 for  these excellent remarks and bibliographical references. Most of them will be included in the revised manuscript. Notice however that the work of Steiner, Attoui, Wimmer& Reischl was already cited.

I agree that further discussion of the aerosol sample flow rate is of clear broad interest, and that the reference to the work of Hontañón and Kruis is most appropriate. Their study achieved the classification of large flow rates of monodisperse nanoparticles through the development of a DMA of heroic dimensions. Their goal was substantially different from that of the Opinion under discussion. Nevertheless, the referee's remark brings the important point that a small hand-held DMA capable of high flow rate of sheath gas should also be able to handle high monodisperse aerosol flow rates, opening up another conceivable application of the DMA development I was trying to encourage. This interesting application would perhaps require somewhat different geometries (wider inlet and outlet slits, etc.). The main lesson following from the referee's insightful remark is that the flow rates may be increased by expanding the DMA dimensions while still running at moderate Reynolds numbers, as done by Hontañón and Kruis.  The same goal may be reached by maintaining conventional moderate dimensions while increasing the Reynolds number well above 2000. This situation is not incompatible with laminar operation in properly designed DMAs, where steady flow has been demonstrated many times with Reynolds numbers well beyond 100,000.

The referee reminds us also of the potential interest of using an electrometer as a detector. The prior comment by Dr. Gerhard Steiner has also noted this fact. In my response to Gerhard’s suggestion I have expanded further the discussion. The referee’s point that the high sample flow rates manageable by electrometers may sometimes achieve greater sensitivities than existing CPCs is indeed nicely illustrated in various instruments developed by Tammet and his colleagues at Tartu (such as the referee's first reference to Asmi et al.). It would be difficult to match the vast sample flow rates of these specialized instruments in a DMA geometry. Nevertheless, it is clear that this opinion should include a related discussion aimed at stimulating the creation of new instruments with resolving powers and sample flow rates intermediate between those of the Tartu analyzers and high flow DMAs.

The referee's reference to Brunelli et al. involves an instrument operating at sheath gas flow rates and resolving powers comparable to those of conventional DMAs. I fully agree with the proposition that the creative geometrical innovations introduced in electrostatic classifiers (and their possible future variants) by Professor Flagan and his students offer great promise for high resolution classification, perhaps even without a need for high flow rates. Nevertheless, it seems to me that this important area of future developments is already discussed in my section 2.7, on the role of geometry.

Thank you again

Juan Fernandez de la Mora