The coma of active comets contains two essential components resulting from cometary activity: gas and dust. To investigate the latter, the Rosetta spacecraft was equipped with several instruments fully dedicated to the analysis of dust in the coma of comet 67P/Churyumov-Gerasimenko (67P). We show that, although not designed to observe dust, another instrument onboard Rosetta can obtain information about dust particles from 67P: the ROSINA-COPS ram gauge (RG, [1]) shown in Figure 1. In particular, it was possible to measure the sublimation of the volatile part of dust agglomerates entering the instrument. From these measurements, we find three different families of volatiles and dimensions, calculated as a diameter of an equivalent sphere of water, of hundreds of nanometres. This value is in accordance with the smallest (refractory) dust structures found so far at 67P. 1. Introduction The main scientific goal of the RG was to measure the ram pressure in the coma of 67P to derive the velocity of the gas. At times, the mostly smooth cometary signal showed sudden increases in measured density (Figure 2). This indicates the presence of a momentary additional source of gas and this is compatible with the sublimation of the volatile component of a dust particle within the RG [2]. We therefore search in the data recorded by the RG for features similar to those in Figure 2, paying attention to exclude those attributable to spacecraft background effects such as offset measurements, slews and thruster firings [3]. COPS had two operating procedures, the so-called monitoring mode and the scientific mode; we studied the measurement datasets of both of them taking into account the strengths and weaknesses of the approaches. The scientific mode provided many measurements of a single feature and allowed monitoring of dust particles that sublimate too fast to be observed by the monitoring mode. However, the science mode has only been used scarcely. On the other hand, the monitoring mode was more frequently active, but operating at lower time-resolution (one minute). Therefore, in most cases, it is not possible to extrapolate information beyond the indication that at that moment there was an agglomerate containing volatiles within the RG. Among all the features, the ones composed of at least five consecutive measurements are investigated (5+ minutes in monitoring mode and 10+ seconds in science mode). Through their distinct decay constants, we analyzed the amount of different groups of volatiles and for the total volume of the volatile part by integrating the obtained signal with time. First, we model the tail of the feature, that is the part that follows immediately after the abrupt increase in density (cf. Figure 2). After subtracting the nominal coma signal, the measurements that make up the tail are fitted by either a single exponential decay function, thus indicating that there is only one volatile component, or by the sum of two different exponentials, thus contemplating the possibility that there are two distinct groups of volatiles. As for the volume of the volatiles, we set a differential equation that describes the variation in the number of volatile particles within the RG as the difference between the sublimating molecules from the agglomerate and the particles escaping from the instrument. Thanks to this differential equation and the previously calculated fit of the tail, we determine the number of molecules of the agglomerate and calculate the diameter of an equivalent sphere based on the assumption that there is only water ice. We opted for this simplistic choice because water is the dominating volatile component in comets [4]. 2. Results We identify the sublimation of 73 agglomerates, 25 of which allow a detailed analysis. Depending on their exponential decay constants, the latter can be divided into three separate families, meaning that there are either three groups of volatiles, or multiple arrangements of the volatiles inside the agglomerates. The volatile sublimation process lasts at most a few tens of minutes, thus explaining why COSIMA [5] detected only refractories: as already pointed out by [6], the interval between collection and analysis by COSIMA is too broad. Moreover, we calculated their size as a diameter of an equivalent sphere of water. We obtain dimensions in the order of hundreds of nanometers, in accordance with the smallest (refractory) dust structures found so far at 67P [7]. 3. Summary and conclusions The RG has proven to be a very versatile tool, since it was able to obtain indications on dust particles, an element for which it was not developed. The data obtained provide redundancy to other instruments of the Rosetta mission, but also add new pieces to the complicated puzzle of cometary activity of 67P and the suggested approach may be an additional tool to better categorize dust agglomerates. References [1] Balsiger H. et al., Space Science Reviews 128, 745-801, 2007 [2] Altwegg K. et al., MNRAS 469, S130-S141, 2017 [3] Schläppi B. et al., Journal of Geophysical Research 115, A12, 2010 [4] Le Roy L. et al., Astronomy & Astrophysics 583, A1, 2015 [5] Fray, N. et al., Nature 538, 72-74, 2016 [6] Altwegg K. et al., Nature Astronomy 4, 533-540, 2020 [7] Mannel T. et al., Astronomy & Astrophysics 630, A26, 2019