Our recent work reported in situ observations of the chemical composition of laboratory-generated aerosol particle surfaces 1. In a Matters Arising (MA) Article we are here replying to, Marchioro et al. raised an issue regarding the system’s detection limit in terms of particle density for the given particle sizes in our measurements. Their means for supporting this claim contains postulates based on a system not identical to ours. Our work 1 does not claim that the vibrational sumfrequency scattering (VSFS) signals are generated explicitly from 40 to 100 nm diameter particles; instead, we stated,“The density of particles was estimated to be ca. 3.8× 106 cm− 3, with a diameter centered at near 40nm, and size distribution spanning from 10 nm to 300 nm.” We feel that due to incompatible comparisons, the MA Article made incorrect assumptions and conclusions about the detection limit of our system in terms of particle density based on the Rayleigh–Gans–Debye (RGD) theory. We believe that the issues raised in the MA Article we here reply to are inapplicable to our work as demonstrated below.(1) Density-dependent non-resonant second-harmonic scattering (SHS) measurements are limited to background Hyper–Rayleigh signals of D2O: the MA Article showed nonresonant SHS with a high-repetition rate laser, which is not relevant to our study. The key difference between their system and ours is the fact that they use a liquid–liquid system while we use a liquid–gas system. Such differences are expected to modify the (1) physical-chemical environment, and (2) the molecular orientation and distribution at the aerosol-ambient interface. For their system, ISHS totalðNÞ ¼ ISHS D2O þ ISHS particleðNÞ, where N is the density of the particles and I is the intensity at the double frequency. Although the SHS signals are linearly dependent on the particle density, the detection limit of non-resonant SHS signals is limited by the background signal from D2O. Thus, the non-resonant SHS of the 100 kHz system in the MA Article does not provide strong evidence against the detection limit of our system.(2) Resonant sum-frequency scattering (SFS) measurements with a 1 kHz laser are limited to low collection efficiency: first, the collection efficiency in the MA Article (1–10 kHz) is at least 10× lower than that in our recent work (100 kHz laser system) 1. In addition, the fluence of the visible pulse plays a vital role in signal intensity: ISFS/IVisIIR, where Ii are the fluences of the relevant beams 2. As this metric was not reported in the MA Article, it is unclear if it was accounted for in the calculations. Further affirmation of the effect of incident fluence is the difference between the IR fluence in our recent work versus those reported in Table 1 of the MA Article, which results in a 16-fold amplification of the SFS signal. Second, the MA Article investigated deuterated hexadecane droplets, with a 109±1nm average hydrodynamic radius, stabilized with 8mM sodium dodecyl sulfate in D2O, with a path length of 200μm. Based on the infrared transmission spectrum acquired for D2O by Carpenter et al. 3, the transmission of the IR pulse energy is< 18% at 2900 cm− 1. The high absorption of D2O and the materials attenuates the IR pulse, resulting in a significant decrease in the generated SFS signals. On the other hand, our system consists simply of water droplets in nitrogen, for which the transmission of all relevant wavelengths involved in the SF process in our system is> 99%. This contrast easily brings signals up by one to two orders of magnitude. Together, the signal levels in our case could be at least 103–104 higher than those in the MA Article. The use of a gas environment not only significantly …