The physics potential of a reactor neutrino experiment with Skipper CCDs: Measuring the weak mixing angle

We analyze in detail the physics potential of an experiment like the one recently proposed by the vIOLETA collaboration: a kilogram-scale Skipper CCD detector deployed 12 meters away from a commercial nuclear reactor core. This experiment would be able to detect coherent elastic neutrino nucleus scattering from reactor neutrinos, capitalizing on the exceptionally low ionization energy threshold of Skipper CCDs. To estimate the physics reach, we elect the measurement of the weak mixing angle as a case study. We choose a realistic benchmark experimental setup and perform variations on this benchmark to understand the role of quenching factor and its systematic uncertainties,background rate and spectral shape, total exposure, and reactor antineutrino flux uncertainty. We take full advantage of the reactor flux measurement of the Daya Bay collaboration to perform a data driven analysis which is, up to a certain extent, independent of the theoretical uncertainties on the reactor antineutrino flux. We show that, under reasonable assumptions, this experimental setup may provide a competitive measurement of the weak mixing angle at few MeV scale with neutrino-nucleus scattering.

We analyze in detail the physics potential of an experiment like the one recently proposed by the vIOLETA collaboration: a kilogram-scale Skipper CCD detector deployed 12 meters away from a commercial nuclear reactor core. This experiment would be able to detect coherent elastic neutrino nucleus scattering from reactor neutrinos, capitalizing on the exceptionally low ionization energy threshold of Skipper CCDs. To estimate the physics reach, we elect the measurement of the weak mixing angle as a case study. We choose a realistic benchmark experimental setup and perform variations on this benchmark to understand the role of quenching factor and its systematic uncertainties, background rate and spectral shape, total exposure, and reactor antineutrino flux uncertainty. We take full advantage of the reactor flux measurement of the Daya Bay collaboration to perform a data driven analysis which is, up to a certain extent, independent of the theoretical uncertainties on the reactor antineutrino flux. We show that, under reasonable assumptions, this experimental setup may provide a competitive measurement of the weak mixing angle at few MeV scale with neutrino-nucleus scattering.

I. INTRODUCTION
The recent discovery of coherent neutrino-nucleus scattering (CEvNS) [1] by the COHERENT experiment [2] has stirred a large deal of interest of the high energy physics community. Although experimentally challenging, the CEvNS cross section opens up a new, exciting venue to detect neutrinos. The key reason lies in the fact that this is a relatively large neutrino scattering cross section. This is because when a neutrino transfers a small amount of momentum to a nucleus, with a de Broglie wavelength comparable to nuclear radii, the neutrino probes the nucleus as a whole, instead of scattering with individual nucleons. Therefore, the CEvNS cross section is sensitive to the square of the nucleus weak charge, Q SM V ≡ N − (4s 2 W − 1)Z, where s W ≡ sin θ W is the weak mixing angle and N, Z denote the number of neutrons and protons in the nucleus, respectively. This is typically referred to as "coherent enhancement" and may enhance the cross section 100-fold or even more, compared to typical quasi-elastic processes. Such a large enhancement opens up the possibility of employing relatively small detectors and still being able to probe interesting and novel physics in the neutrino sector.
Nevertheless, observing low energy nuclear recoils is still a challenge overcome by only few detectors. Among those are high purity germanium detectors, as used for example in the CoGeNT [3] and GEMMA [4] experiments, and dual phase xenon time projection chambers, such as those used in XENON1T [5]. A more recent and promising technology, and the main focus of the present paper, are Skipper charge couple devices (Skipper CCDs).
Skipper CCDs are 2D pixelated detectors fabricated on high resistivity silicon [6]. Its small pixel size of 15 µm by 15 µm allows for good spatial resolution and event discrimination. Devices with several millions of pixels can be fabricated with high yield. The high resistive substrate permits an active depth of each pixel of 675 µm totaling up to approximately 8 grams of active silicon per device. In particular, the great advantage of the Skipper CCD is the high energy resolution thanks to its very low readout noise. It provides single-charge counting capability for any ionization packet in the active volume [7,8]. The energy resolution is then limited by the silicon absorption protons. Neutral current interactions can be parametrized by the following effective Lagrangian Here, G F is the Fermi constant, τ 3 denotes the third Pauli matrix, sin 2 θ W ≡ 1 − M 2 W /M 2 Z is the weak mixing angle (M W,Z are the masses of the weak gauge bosons), Q f is the electric charge of the fermions f and the sum runs over f = ν, u, d. In this case, the differential standard CEvNS cross section can be written as where m N and E R are the mass and recoil energy of the struck nucleus and E ν is the incoming neutrino energy. As mentioned in the introduction, the weak charge of the nucleus is defined as where N and Z are the number of neutrons and protons in the target nucleus and F(E R ) is a form factor that parametrizes the coherence of the interaction. It is important to notice that this cross section is proportional to the square of the number of constituents, such as expected in a coherent scattering, with the modification introduced by the 1 − 4 sin 2 θ W factor, coming from the different effective weak coupling of protons and neutrons. The form factor describes the coherence of the interaction. Here, we will focus on the simplistic Helm form factor [39,70,71], where k ≡ √ 2m N E R , s 1 fm is the nuclear skin thickness, r = √ R 2 − 5s 2 , and R 1.14 (Z + N ) 1/3 is the effective nuclear radius. This form factor parametrizes the internal structure of the nucleus as seen by the incoming neutrino.
To have a better understanding of all those quantities, it is useful to plug in numbers relevant to our study into these formulas. We will study the weak mixing angle sensitivity of an experimental setup consisting of a Skipper CCD detector deployed nearby a commercial nuclear reactor. For typical neutrino energies at the MeV scale, the maximum nuclear recoil achievable for a nucleus N is where m p 0.938 GeV is the proton mass. Here we can see the difficulty in detecting CEvNS with reactor neutrinos: the detector needs to be sensitive to energy depositions of 100 eV or so, to be able to take advantage of the large neutrino flux. Conversely, the minimum neutrino energy for a given nuclear recoil is Finally, the typical momentum transfer q for the recoil energies of interest is The form factor is more complicated to write in a useful form. Nevertheless, noticing that allows us to expand the form factor for small momentum transfer yielding F ∼ 1 + O(k 2 s 2 , k 2 r 2 ). Therefore, for this experimental setup, the form factor is very close to F(E R ) ∼ 1 and we will approximate it to unity henceforth. Using this approximation, we can approximate the differential cross section by  Note however that Q SM V depends on the weak mixing angle. The last expression we will present here draws upon the fact that sin 2 θ W 0.238 is close to 1/4. By writing Q SM For example, from the above formula we expect a 1% change in sin 2 θ W to correspond roughly to a 2% change in the cross section, see Fig. 1.
Finally, a few words regarding the weak mixing angle. The definition of the weak mixing angle and its renormalization group running depends on the renormalization scheme. In this work we adopt the MS renormalization scheme, in which the weak mixing angle for very low momentum transfer is predicted to be sin 2 θ W (Q 2 ) = 0.23867 from the measurement of the same angle at the Z boson pole [72]. In the standard model, sin 2 θ W does not run considerably below Q 2 ≡ −q 2 ∼ (100 MeV) 2 or so. Therefore, we will assume this value in the following as input in order to estimate the sensitivity of our experimental setup to the weak mixing.

III. ANALYSIS
The experimental setup we consider here consists of a Skipper CCD detector, deployed at 12 meters from the main core of a nuclear reactor like the Atucha II reactor complex, located at the province of Buenos Aires, Argentina. Throughout our analysis, the detector is assumed to have a fiducial mass of 1 kg. Later, we will comment on the physics case of a 10 kg detector. The core is assumed to have a thermal power of 2 GW, emitting about ∼ 10 28 electron antineutrinos per year. We will assume a data taking period of 3 years.
Regarding backgrounds, it is expected that the Compton interactions from high energy photons and nuclear interactions from high energy neutrons will be the dominant contributions for the recoil energies of interest. Both processes are well predicted by state-of-the-art particle simulators, but should be analyzed for a specific passive shield configuration to have their impact properly quantified. Therefore, this experimental setup depends crucially on background measurements during reactor-off periods, which are made as short as possible in commercial nuclear reactors. To be realistic, we assume an average reactor-off time of 45 days per year. For the background estimation, we call attention to the measured background in the CONNIE experiment of about 10 kdru (1 differential rate unit, or dru, is 1 event per day-keV-kg) [73]. New studies on the CCD used by CONNIE show that some of its background can be produced by a partial charge collection layer in the back of the sensor [74]. Large charge packets created in this highly doped region are observed as lower energy events since some of the carriers are lost by recombination. This problem can be eliminated by a back side processing of the sensor as explained and tested in the same publication. The CCD detector of CONNIE was located right outside the reactor dome with almost no virtual overburden besides the passive shield. In the scenario considered here, the detector is inside the dome providing some extra reduction to cosmic background. Therefore, we will assume a baseline background of 1 kdru, with flat deposited energy spectrum. We will study the impact of higher or lower background rates, as well as the impact of the background spectral shape [75]. Moreover, we will consider the reactor-on background to be negligible, although this needs to be determined by detailed simulations and mandatorily avoided by adding shielding between the detector and the reactor core.
Skipper CCD sensors have shown very low leakage current contributions [76,77] that allow to explore ionizations up to one or two electrons. Nevertheless we assume a conservative minimum of ionization energy of 15 eV, corresponding on average to four ionized electrons. It should be noted that the translation from nuclear recoil energy to ionization energy, which is what is detected by Skipper CCDs, is encoded by the quenching factor. At such low recoils, the quenching factor is not well-known, and measurements to determine it are planned using the Skipper sensors following the procedure presented in Ref. [78]. This is critical in reactor neutrino CEvNS, as the neutrino flux grows considerably at lower neutrino energies. We will estimate the effect of two different quenching functions and its uncertainties on the experimental sensitivity. Our benchmark will assume the parametrization from Ref. [18] of the measurement performed in Ref. [78]. As the expected neutrino ionization energy spectrum is fairly broad, and the reconstruction of the ionization energy resolution in Skipper CCDs is excellent (essentially only affected by the silicon absorption, which can be estimated through the fano factor [9]), we do not consider any ionization energy smearing here. Note however that this does not help in reconstructing the incoming neutrino energy. We bin the simulated data in 50 eV ionization energy bins.
Finally, several systematic uncertainties may affect the physics reach of these experiments. Among the most important ones is the uncertainty on the reactor neutrino flux. Several theoretical estimates of this flux have been performed [79,80]. Although the calculation of these fluxes involve thousands of beta branches and forbidden decays, they achieve a remarkable systematic error of about 2%. Nevertheless, these calculations do not agree with the measured reactor antineutrino flux via the inverse beta decay process, giving rise to a 3σ anomaly, dubbed the reactor antineutrino anomaly. The reason for this discrepancy is still unknown, ranging from too aggressive uncertainties [81][82][83] to new physics beyond the standard model [84][85][86]. On top of that, theoretical calculations fail to predict a relatively large feature in the reactor neutrino spectrum around neutrino energies of 5 MeV (see e.g. Ref. [87]). In view of that, we will estimate the experimental sensitivity to sin 2 θ W using the flux determination from the Daya Bay experiment [69] instead of relying on theoretical calculations of the neutrino flux and the corresponding systematic uncertainties. Although the former has larger uncertainties, it is a data-driven approach, and thus it is more robust than theoretical estimates. Later, we will show that the flux uncertainty does not play a role as important as other factors such as background rate and quenching factor determination.
The expected signal event rate, for each fuel isotope, measured by the Skipper CCD detector can be obtained by convolving the neutrino flux with the CEvNS cross section, taking into account the quenching factor, namely, We denote the ionization energy by E I , and the nuclear recoil energy by E R . W is the total reactor power. For convenience, we have assigned three indices to n q aj . The upper index q indicates the fuel isotope, that is, 235 U, 239 Pu, 238 U and 241 Pu. Associated to that is the relative rate per fission f q and energy released per fission e q for isotope q . The first lower index corresponds to the ionization energy bin, while the second one refers to the incoming neutrino energy integration interval. Including the latter is important as it allows us to properly take care of flux uncertainties. The quenching function is Q(E I ) = E I /E R . Note that the terms in the outer integral come from the Jacobian when changing the integration variable from E R to E I . The intervals of integration should be appropriate for the binning and also respect Eq. (5). The neutrino flux per fission for each isotope is given by φ q νe . For energies above the inverse beta decay threshold, E ν > 1.8 MeV, we use the antineutrino fluxes measured by Daya Bay [69]. In their analysis, Daya Bay uses the theoretical estimates from Refs. [79,80] as inputs for 238 U and 241 Pu and extracts the fluxes for the two isotopes with the largest relative rates per fission, 235 U and 239 Pu, from their measurements. For energies lower than the inverse beta decay threshold we will use the flux estimate of Vogel and Engel [88]. We will assume an uncertainty of 5% in the relative rate per fission of the flux for every isotope.
The sensitivity over sin 2 θ W is determined by minimizing the following χ 2 for the reactor-on running, There are several quantities in this expression that need to be defined. Let us start with the last four terms which correspond to the Gaussian distributions (also sometimes referred as Gaussian priors) that parametrize the systematic uncertainties of the present experiment. The first systematic is related to the background determination. During reactor-off running, the experiment will determine the background in each bin. The relative statistical uncertainty of this determination is what enters in σ B a . The second systematic is on the total reactor power W . The third systematic corresponds to the relative rate per fission f q , which depends on the isotope composition. The last one is on the quenching factor Q(E I ). The α's are pull parameters that will be minimized for each value of sin 2 θ W according to the profiling method. The approach adopted here, regarding the determination of the background, is most conservative. In principle, one can also adopt a more aggressive approach. For example, if the background can be well modeled by MonteCarlo simulations, say by a polynomial function, the reactor-off running would determine the uncertainties on the coefficients of such function, which would then be used in Eq. (11) (the implementation of the pull parameters would change accordingly). Now we move on to the first term of the χ 2 . Here, d and t are vectors of data (Asimov data with sin 2 θ W = 0.238) and theory (where sin 2 θ W is allowed to vary) events, respectively. These vectors can be constructed by summing up all neutrino energies and isotope contributions in the signal rate in Eq. (10), as well as background rates.
where B a is the background in ionization energy bin a. Note that the theory prediction, t, should also properly include the pull parameters α. To do so, the following substitutions should be made Note that the relative rate per fission should sum up to 1, and therefore q (1 + α q )f q = 1.
where systematic and statistical uncertainties are encoded in the first and second terms, respectively. Note that the Daya Bay collaboration only provides this correlation matrix for the isotopes with the largest rate per fission, namely, 235 U and 239 Pu, and for neutrino energies above 1.8 MeV (which is the threshold for inverse beta decay). In this case, V q ij are taken from Daya Bay's flux covariance matrix [69] and correlate the uncertainties in the measured spectrum. For the flux below 1.8 MeV, we use a single neutrino energy bin, and we assume an uncorrelated spectral uncertainty of 5%. Moreover, for the subdominant rate per fission of the isotopes 238 U and 241 Pu, we use a 5% bin-to-bin uncorrelated systematic uncertainty on the flux. Given that the relative rate per fission of these isotopes are assumed to be 0.07 and 0.06, such systematic uncertainties play a small role in the analysis.
The numerical values for the priors used can be found in Table. I. Note that the background uncertainty is derived from the reactor-off period. We assume the background rate to be 1 kdru, flat in ionization energy, an exposure of 45 days per year, for 3 years, and a detector mass of 1 kg. This yields a statistical determination of the background rate of 1.2% in each 50 eV ionization energy bin.

IV. RESULTS
We now proceed to the results of our simulation. Our benchmark setup is summarized in Table II. We assume the quenching factor to be the one measured in Ref. [78] and parametrized in Ref. [18]. The quenching models and the impact of systematic uncertainties will be discussed in detail below. The signal and background event rates spectra for this benchmark are presented in the left panel of Fig. 2. We will perform studies on variations of this scenario to understand the role of different aspects on the experimental sensitivity. The sensitivity to the weak mixing angle in our benchmark scenario is shown in the right panel of Fig. 2 for two different assumptions on the quenching factor which will be discussed below, the Lindhard and the Chavarria models, without systematic uncertainties on the quenching. Together with our estimate, we also present current measurements (taking into account the running of sin 2 θ W ) [72] and a forecast for the measurement on the DUNE near detector complex [89]. The precision achieved by this benchmark setup is 1.4% for Lindhard and 2.8% for Chavarria quenching, at mean momentum transfers Q 2 of 4.3 MeV and 6.6 MeV, respectively. The mean momentum transfer is different because the minimum detectable recoil energy depends on the quenching. As we can see, under reasonable assumptions, this experimental setup would determine the weak mixing angle with a precision similar or slightly worse than atomic parity violation (APV), and comparable precision to several other determinations, including the future DUNE experiment [89]. The competitiveness of such a small experiment is quite remarkable. The measurement proposed here would be one of the very few determinations of the weak mixing angle with neutrinos, which is particularly important given the discrepancy between measurement and theory prediction observed in the NuTeV experiment [90]. As a side comment, by using the LEP determination of the weak mixing angle and the theoretical prediction for the reactor antineutrino flux, this setup could probe the reactor anomaly at the 3 − 6% precision. To estimate how robust this measurement is and how it depends on our assumptions, we proceed to a study of how each ingredient affects our analysis. We will study the role of quenching factor parametrization and uncertainties, as well as background rate and shape.  Table II). The vertical dashed line shows the ionization energy threshold of Skipper CCDs. Right: Sensitivity of our experimental setup to sin 2 θW compared with different experiments [72,89] in the MS renormalization scheme. The two colored data points correspond to two different quenching factors, Lindhard (green) and Chavarria (dark blue). No uncertainty on the quenching was considered.

A. Quenching factor
As mentioned before, the quenching factor characterizes the fraction of the recoil energy transformed into observable ionization energy. Measurements of the quenching factor in the case of silicon have been previously performed [78]. Low energy measurements disagree with the theoretical prediction obtained using the Lindhard model [91] for the quenching. Given this discrepancy, we will study two cases for the quenching factor. Lindhard model. We start by describing the implementation of the widely used Lindhard model for the quenching factor at low energies. We perform a polynomial fit to the Lindhard quenching model given in Ref. [92], obtaining the following functional form  5), the minimum neutrino energy to produce such recoil is about 0.9 MeV. Therefore, assuming the Lindhard quenching model, a large portion of the reactor antineutrino flux, below the inverse beta decay threshold, can lead to observable recoils in our experimental setup.
Chavarria model. The measurements performed to determine the quenching factor at low energies [78] can be parametrized by the following ratio of polynomials A fit to the data determines p 0 = 56 keV 3 , p 1 = 1096 keV 2 , p 2 = 382 keV, p 3 = 168 keV 2 and p 4 = 155 keV. This parametrization, which we will refer to as the Chavarria model, has been used by the CONNIE experiment in beyond standard model physics searches [18]. Given that the quenching factor has only been measured down to ionization energies of about 60 eV and the measurement is not in agreement with the theoretical predictions of Lindhard, it is not completely clear what should be assumed for the quenching below that energy. Thus, we use this parametrization as an alternative to the Lindhard quenching factor described above. One important feature of the Chavarria quenching is that, for small enough E I , the recoil energy E R = E I /Q becomes constant. Therefore, it becomes very hard to detect low recoils, as the ionization energy shrinks very quickly for small E R . Plugging in numbers, the 15 eV ionization energy threshold translates into a recoil energy of 425 eV, which in turn gives us a minimum neutrino energy of 2.3 MeV, see Eq. (5). Thus, compared to the Lindhard quenching, the Chavarria quenching will significantly decrease statistics. Besides, a systematic uncertainty on the Chavarria quenching will change the minimum neutrino energy considerably and therefore significantly affect the rate of signal events. We emphasize however that this assumption about the quenching is very conservative and future measurements will clarify the quenching below 60 eV of ionization energy.
Quenching factor systematics. To understand the impact of systematic uncertainties we evaluate the sin 2 θ W sensitivity under several assumptions and present the results in Fig. 3. The assumptions regarding quenching and its systematic uncertainties are the following: Lindhard (green) or Chavarria (dark blue) models, with 0% (solid), 10% (dotted) and 25% (dashed) overall normalization uncertainties in the quenching (note that for Lindhard these lines are all on top of each other). As discussed, the Lindhard quenching yields a considerably superior result compared to the Chavarria model due to statistics. Moreover, in the Lindhard model a variation in the quenching does not change the rate of signal events by much, due to the very low thresholds of Skipper CCD detectors. For any of the quenching uncertainties, the Lindhard quenching would allow for a 1.4% precision measurement of the weak mixing angle. For the more conservative quenching model of Chavarria, we find that the precision on the mixing angle is significantly decreased to be 2.8%, 4.2% and 6.0% for a 0%, 10% and 25% systematic uncertainty, respectively. This dramatic effect is simply due to the fact that variations in the quenching lead to significant variations in the signal rates, as much more or much less neutrino flux becomes detectable. We conclude, therefore, that the determination of the quenching factor at low ionization energies is crucial for the physics case of an experiment measuring CEvNS from reactor neutrinos with a Skipper CCD detector.

B. Backgrounds
In CEvNS, unless the recoiling nucleus is tracked and reconstructed, which is extremely challenging, the typical experimental signature is quite simple: just a small deposition of energy in the detector. Because of that, background rejection is a difficult task in CEvNS detectors. This is particularly true for Skipper CCDs. Due to their slow readout, not even active vetos may be used to reject the cosmic induced background. Background rates and measurements are therefore crucial to the success of the experimental setup considered here. We will quantify this statement in the following by analyzing the dependence of the experimental sensitivity on the background rate, modeling, and shape. Our benchmark assumes reactor-off running time of 45 days per year which serves to determine the background rate. In the background determination, we simply use the uncorrelated bin-to-bin statistical uncertainty from the reactoroff data taking period. A background model could significantly improve the experimental sensitivity, but that would require a dedicated background study, which is beyond the scope of this paper. For concreteness, we will use the Chavarria quenching as it will result in a more conservative sensitivity estimate, but without quenching systematic uncertainty.
First, we investigate the effect of increasing or decreasing the background rate. In the left panel of Fig.4, we present relative 1σ determination of sin 2 θ W as a function of the total background rate for both Chavarria (blue line) and Lindhard (green line) quenching factors. We observe that, in the case of Chavarria quenching, the measurement of the weak mixing angle is highly sensitive on the total background. For instance, if the background is 10 kdru, that is, ten times larger than in our benchmark scenario, the precision on sin 2 θ W gets worse from 2.8% to 8%, while for 0.1 kdru background rate the precision is enhanced to 1.6%. The dependence on the total background rate is less pronounced for Lindhard quenching, as the signal rate in this case is much higher, and therefore the experimental sensitivity is less affected by backgrounds around 1 kdru.
To see how the shape of the background changes the sensitivity, in the right panel of Fig. 4, we perform the analyses for three representative spectral shapes in ionization energy: flat (black), proportional to E −1 I (dashed dark cyan), and linear in E I (dotted orange). In doing these studies, we have kept the background rate fixed at 1 kdru in the energy range E I ∈ [15, 675] eV. As already pointed out in Ref. [75], if the background grows at low energies, the experimental sensitivity is significantly decreased. This is because in that case, the background has a shape very similar to the signal, see Fig. 2. Conversely, if the background shrinks at low ionization energy, the experimental sensitivity would be increased. From this, we can appreciate the relevance of properly modeling and mitigating the background rate in experiments employing Skipper CCD detectors.

C. Detector mass
We also perform a study the dependence of the sin 2 θ W determination as a function of the total exposure. This helps to understand if the sensitivity is systematic or statistically limited. In Fig. 5, we show, for our benchmark scenario, how the 1σ uncertainty on sin 2 θ W depends on the total exposure for Chavarria (blue line) and Lindhard (green line) quenching. In the first case, going from 3 kg-year to 30 kg-years of exposure, the precision on the weak mixing determination could go from 2.8% to 0.9% evidencing that the experiment sensitivity is not systematics limited, and thus a larger detector or longer exposure time would yield significantly better physics measurements. We observe a similar pattern for the Lindhard quenching case.

D. Reactor antineutrino flux uncertainty
As mentioned above, in view of the recent reactor antineutrino anomaly, one may find more robust to do precision analysis with the Daya Bay flux determination. Nevertheless, it is a useful exercise to see the dependence of the  sensitivity to sin 2 θ W on the reactor antineutrino flux uncertainty. To do so, we perform our last analysis in this manuscript under three assumptions regarding the flux uncertainties, in Fig. 6: Daya Bay covariance matrix [69] (solid blue), 2% theory normalization error [79,80] (dotted blue) and 5% flux normalization error [81][82][83] (dashed blue). As we can see, the Daya Bay flux determination is already quite good for our benchmark sensitivity. By reducing the flux uncertainty from Daya Bay's covariance matrix to a 2% overall normalization, the sensitivity improves marginally. We therefore conclude that, although an improved flux model or determination would be beneficial, this is not a bottleneck in our experimental proposal.

V. CONCLUSIONS
In this paper, we have analyzed the sensitivity to the weak mixing angle of an experimental configuration like the recently proposed vIOLETA experiment, a Skipper CCD detector deployed 12 meters away from the core of the commercial nuclear reactor Atucha II in the province of Buenos Aires, Argentina. We have analyzed the impact of several experimental aspects on the estimated sensitivity: quenching factor and its uncertainty, background rate and spectral shape, and total exposure. We have quantified the crucial role of the quenching factor on the sensitivity. Nevertheless, the role of systematic uncertainties in the quenching depends on the quenching itself. For the Lindhard quenching model, these systematics play a small role, while the opposite happens for the conservative Chavarria parametrization. As expected, the background rate is also critical to the success of a neutrino experiment leveraging the Skipper CCD technology. A flat background of 1 kdru would allow for a competitive measurement of the weak mixing angle, while a rate of 10 kdru would be prohibitive. Moreover, if the background follows a E −1 I spectral shape, similar to the signal, mitigation becomes of the utmost importance. Finally, our findings also show that the experimental sensitivity is not systematics limited for a 3 kg-year exposure. Therefore, better measurements may be achieved by deploying a larger detector. Our findings show that, under realistic assumptions, the measurement of the weak mixing angle performed by this experimental setup would be competitive with other experiments, including DUNE. This measurement would be one of the very few competitive determinations of the weak mixing angle using neutrinos, which is particularly important given the unsettled NuTeV discrepancy. We hope our results can be useful to experimentalists as a prioritization guide for maximizing the physics output of a reactor neutrino Skipper CCD experiment.