*3.5. Reflectance*

The measured reflectance spectrum of the thickest uncapped CuCl film (Q10 1000 c.) is shown in Figure 10a. It shows a significant increase of the reflected intensity compared to the bare substrate, which is indicative of a larger optical density of the film. In addition, a fairly strong spectral structure with maximum intensity at about 3.3 eV has apparently at least two narrow components. The latter are better seen in the twice differentiated spectra plotted in Figure 10b. The differentiation enhances the sharper structures and suppresses the flat background. Based on the narrow structures in the derivative spectra, we have fitted the measured reflectance with a model consisting of a flat background dielectric function, with two superposed Gaussian absorption bands. The model lineshape of Figure 10a,b represents the best-fit results with the (fitted) value of the film thickness of 29 nm, and the background value of the real part of its dielectric function of 3.2. The resulting parameters of the bands are listed in Table 7. The strength parameter is proportional to the area below the absorptive (imaginary) part of the dielectric function, having the units of eV; however, we are interested in relative values and use arbitrary units here.

**Figure 10.** (**a**) Normal incidence reflectance spectra of the uncapped, 1000 c. sample; the positions of two prominent Gaussian bands are indicated by arrows, bare substrate (dashed line); (**b**) second derivative; and (**c**) second derivative of the reflectance spectrum from the capped, 500 c. sample. Measured data (symbols), model lineshapes (red solid and dashed lines).



The two Gaussian bands are 0.08 eV apart, which is in a good agreement with the distance of 0.06 eV seen in the absorption spectra of *Z*<sup>3</sup> (3.23 eV) and *Z*1,2 (3.308 eV) excitons at in the PL results and those seen at 4 K [44]. In addition, the significantly larger strength of the upper (3.30 eV) band is clearly seen in the reflectance results, in contrast to the PL intensities. This is similar to the relative intensities seen in direct optical absorption measurements [18,45]. We have also observed a weak and broad luminescence band centred at about 2.7 eV, with the maximum signal below 4% of the main PL band; its high energy tail forms the smooth background seen in Figure 7. This may be due to radiative recombination of defect states in the CuCl crystallites, as well as other species present in the deposited layer.

A film of the sample Q8 (500 c. with an Al2O3 capping layer) displays a reduced intensity excitonic structure in the differentiated reflectance, see Figure 10c. The results are compatible with reduced effective film thickness of approximately 12 nm. In addition, the upper lineshape seems to be narrower than that of the thicker sample Q10, probably due to the smaller dispersion of the crystallite sizes. It can be seen that reflectance measurements are a useful way to identify the presence of CuCl in thin films. This method would be particularly useful where the film is deposited on a non-UV transparent substrate.

The reflectance of sample Q8 was remeasured after approximately three weeks in normal atmosphere. A comparison between the two measurements is shown in Figure 11. There is no significant change to the excitonic absorption confirming the XPS results that 5 nm of ALD Al2O3 provides an effective protection layer for the CuCl. This is consistent with published work showing that ALD Al2O3 constitutes a good moisture diffusion barrier when applied to other materials [46].

**Figure 11.** Comparison of reflectance curves (2nd derivative) for aluminium oxide-capped sample Q8 before and after approximately 3-week exposure to ambient air.

#### **4. Conclusions**

It has been demonstrated that CuCl can be deposited by sequentially pulsed vapour deposition process using the precursors [Bis(trimethylsilyl)acetylene](hexafluoroacetylacetonato)copper(I) and pyridine hydrochloride and that an in situ ALD capping layer of aluminium oxide is an effective barrier for preventing atmospheric degradation. The crystal structure has been shown to be the zinc blende-structure γ-CuCl by XRD. The crystallites become more facetted as the film thickness increases. The bulk of films shows only some organic contamination from incomplete precursor reaction, however, there is heavier organic contamination on the surface, again from unreacted precursor molecules. The characteristic photoluminescence behaviour of CuCl has been shown with emissions from the *Z*1,2, *Z*<sup>3</sup> and bound excitons. The chemical composition was investigated by XPS: the 200 c. and 500 c. films show surface layers with predominantly Cu+ bonding from CuCl with some organic contamination. The thicker films with 1000 deposition cycles have significant Cu2+ content from CuF2 or F containing organic fragments. Optical reflectance measurements have shown that the characteristic exciton absorptions can be detected, at similar energies to optical absorption features measured by transmission. This enables exciton absorption measurements to be carried out on nontransparent samples.

The overall results show that this is a method for deposition of nanocrystalline arrays suitable for further investigation for the development of new optoelectronic structures and devices. Further studies are also required to clarify the difference in chemical composition between the surface layers and the bulk of the film.

**Author Contributions:** Conceptualization, D.C.C; Methodology, D.C.C.; Formal Analysis, T.H., R.K., and D.C.C.; Investigation, R.K., T.H., O.C., K.K., R.Z., J.P., and J.M.M.; Resources, R.Z., J.P., and J.M.M.; Writing–Original Draft Preparation, T.H. and J.H.; Writing–Review & Editing, D.C.C., J.H., R.Z., J.M.M., R.K., T.H., K.K., and O.C.; Visualization, D.C.C. and J.H.; Supervision, D.C.C., J.H., and J.M.M.; Project Administration, D.C.C. and R.K.; Funding Acquisition, D.C.C. and J.M.M.

**Funding:** This work was financially supported by the Czech Science Foundation (No. 17-02328S), the Ministry of Youth, Education and Sports of the Czech Republic (Nos. LM2015082, LQ1601, LO1411 (NPU I), and CZ.02.1.01/0.0/0.0/16\_013/0001829), and the European Regional Development Fund (project CZ.1.05/2.1.00/03.0086).

**Acknowledgments:** The authors acknowledge the support of COST action MP1402 HERALD in the course of this work.

**Conflicts of Interest:** The authors declare no conflicts of interest.
