Bridging energy bands to the crystalline and 2 amorphous states of Si QDs

The relationship between crystallization process and opto-electronic properties of silicon quantum 16 dots (Si QDs) synthesized by atmospheric pressure plasmas (APPs) is studied in this work. The synthesis of Si QDs is carried out flowing silane as gas precursor in a plasma confined to a submillimeter space. Experimental conditions are adjusted to propitiate the crystallization of the Si QDs and produce QDs with both an amorphous and a crystalline character. In all the cases, Si 20 QDs present a well-defined particle mean size in the range of 1.5-5.5 nm. Si QDs present optical bandgaps between 2.3 eV and 2.5 eV, which are affected by quantum confinement. Plasma 22 parameters evaluated using optical emission spectroscopy are then used as inputs for a collisional plasma model, whose calculations yield the surface temperature of the Si QDs within the plasma, justifying the crystallization behavior for certain experimental conditions. We measure the ultraviolet-visible optical properties and electronic properties through various techniques, build an energy level diagram for the valence electrons region as a function of the crystallinity of the QDs and finally discuss the integration of these as active layers of all-inorganic solar cells.


45
Atmospheric pressure plasmas (APPs) present great versatility for the production and 46 treatment of nanomaterials [24,25] for they allow flexible design and easy integrability. Also, at this pressure, ion collisions with the surface of nanoparticles are responsible for particle heating 48 above the background gas temperature, allowing a controlled crystallization [14] by carefully 49 tuning the synthesis conditions. We have previously studied the synthesis and materials properties 50 of crystalline and amorphous silicon QDs, separately, by atmospheric pressure plasmas (APPs)

53
Herein, we present an experimental and theoretical investigation of the Si QDs phase transition 54 in APPs, comparing the plasma conditions leading to or preventing crystallization.

55
We then perform various measurements on selected samples to assess the energy band diagrams 56 and derive the relationships between structural features and opto-electronic properties in function 57 of synthesis conditions. In this context we use different measurements techniques to build an 58 energy level diagram of near-gap electron states, and critically compare methods and results. This 59 approach is important for implementing nanomaterials in real world applications.

60
Finally, we test the applicability of our Si QDs integrating them as active layers in all-inorganic 61 solar cells. While these devices still present very low efficiencies, here we demonstrate the 62 viability of APP processes to be used in the manufacturing of next generation photovoltaics.

Experimental details 65
The plasma reactor used for the synthesis of Si QDs operates in a parallel electrode 66 configuration at atmospheric pressure (760 Torr). A schematic diagram of the system is depicted 67 in Figure S1 of supporting information. The plasma is generated inside a rectangular glass tube 68 with a 0.5 mm gap and 0.3 mm of wall thickness. Radio frequency (RF) power at 13.56 MHz and 69 120 W is applied through a matching unit and to two rectangular copper electrodes with a section 70 of 20 mm x 5 mm.

71
Argon and hydrogen are supplied as background gases, while silane (SiH 4 ) is used as Si

127
The spots detected in the diffractogram match well with crystalline planes corresponding to the 128 diamond lattice of silicon (see Figure S2 of the supporting information). On the other hand, a high  hydrogenation therefore seems to be important both in the oxidation process as well as at some 199 level determining the phase of the QDs.

200
The XPS instrument also allows the acquisition of reflection electron energy loss spectra using 201 the flood gun as an electron source. This technique can be used to easily ascertain qualitatively 202 the presence of incorporated hydrogen within a sample, via an energy loss feature which sits 203 around 1.8 eV from the zero-loss peak. In our case we observe a ( Figure S4, supporting       Table 1). The VBM 267 is calculated also from the UPS measurements where VBM UPS = E F-UPS -E on-set (Table 1).

268
Fermi level values obtained with the Kelvin Probe show the same trend as the ones obtained by 269 UPS, even though the values are different, and in one case (50 ppm) particularly higher.

270
Meanwhile XPS produces the difference between the Fermi level and the VBM (Figure 4b)

280
All the values summarized in Table 1 are also reported in Figure 5. The comparison shows that 281 discrepancies from different measurements do exist, which in most cases are within measurement

305
Instead XPS in the valence band region suffers from moderate energy resolution, lower yields and 306 the absence of an energy cut-off which does not permit to obtain values relative to the vacuum 307 level and generally scopes a "deeper" region beneath the surface (5-10 nm) [34]. Both techniques 308 though are operated in high vacuum, which can help preserving the state of surfaces.

309
KP and associated APS operated in atmosphere are instead subject to surface dipoles and electrostatic potential at the atmospheric conditions at the time of measurements [35].  Table 1) and as previously observed ( Figure 1e behavior. This implies that charge transport within the layer of amorphous particles may be 365 superior in respect to the films of crystalline nanoparticles [43].

375
We believe in conclusion that UPS values are more reliable both for the determination of

385
In this section we finally provide some details on the integration of Si QDs in photovoltaic (PV)

386
devices. This is mainly to show that a better knowledge of the EBD parameters can benefit the 387 application development and also in part to demonstrate the capability of the APP process for Supporting Information).

394
One of the device architectures is illustrated in figure 6a, the other in supporting information 395 S10. then used as an electron blocking layer and a sputter-deposited gold film is used as a top contact.

402
The non-equilibrated band diagram of the cell is shown in figure 6. We should note that the contacts; however, the current density is very poor and is affecting the overall performance.

409
Additionally, the measured value of series resistance is also indicative of relatively efficient 410 electron transport in the cell (table S10 in supporting information). Our current set-up could not