Zn-doped P-type InAs Nanocrystal Quantum Dots

Doped heavy metal-free III-V semiconductor nanocrystal quantum dots are of great interest both from the fundamental aspects of doping in highly confined structures, and from the applicative side of utilizing such building blocks in the fabrication of p-n homojunction devices. InAs nanocrystals, that are of particular relevance for short wave IR detection and emission applications, manifest heavy n-type character poising a challenge for their transition to p-type behavior. We present p-type doping of InAs nanocrystals with Zn-enabling control over the charge carrier type in InAs QDs field effect transistors. The post-synthesis doping reaction mechanism is studied for Zn precursors with varying reactivity. Successful p-type doping was achieved by the more reactive precursor, diethylzinc. Substitutional doping by Zn2+ replacing In3+ is established by X-ray absorption spectroscopy analysis. Furthermore, enhanced near IR photoluminescence is observed due to surface passivation by Zn as indicated from elemental mapping utilizing high resolution electron microscopy corroborated by X-ray photoelectron spectroscopy study. The demonstrated ability to control the carrier type, along with the improved emission characteristics, paves the way towards fabrication of optoelectronic devices active in the short wave IR region utilizing heavy-metal free nanocrystal building blocks.


Introduction
The field of semiconductor colloidal quantum dots (CQDs)-based optoelectronics draws attention worldwide, as new technologies call for high volume, cost-effective fabrication of optoelectronic devices with precise control over the optical properties and complementary metal-oxide-semiconductor (CMOS) compatibility. 1,2 Of specific interest are QDs with tunable band gap across the near infrared and short-wave infrared (SWIR) spectra for applications in IR detection and vision 3,4 , photovoltaics, [5][6][7] and telecommunications. [8][9][10][11] Numerous efforts and focus have been given to Pb-and Hg-based semiconductor QDs, owing to the well-established synthetic procedures and well-studied optical properties, perfected over the years. [12][13][14][15][16] However, the implementation of such materials in commercial optoelectronics is restricted due to environmental concerns and regulatory aspects as dictated by the Restriction of Hazardous Substances Directive (RoHS). 17 This stimulates the research and development of novel alternatives, where III-V colloidal QDs such as InAs are considered as a promising candidate material system for infrared optoelectronic applications. 8,18 A particular challenge in InAs QDs, and also for other colloidal QDs, is the control and understanding of doping, and its utilization in optoelectronic devices. [19][20][21][22][23] The study of III-V colloidal QDs based optoelectronics, and InAs in particular, is hindered, mainly due to synthetic issues, including precursors toxicity and limited control over the size and uniformity of the QDs. In the last few years, several synthetic advancements presented improved tunability of the band gap across the NIR and SWIR. [24][25][26][27] Controlled growth of InAs nanocrystals by continuously supplying InAs nanoclusters, enables the synthesis of highly crystalline and monodisperse nanocrystals with narrow absorption features across the near infrared. 28,29 Additional successive growth cycles provide the ability to grow even larger InAs nanocrystals, pushing the absorption feature towards 1600 nm while maintaining a narrow size distribution. 30 InAs QDs, as synthesized, present n-type character. 22,31 Thus, combining such n-type InAs QDs with another p-type material, allowed to form a heterojunction as was achieved for photovoltaic 29 and IR detection applications. [32][33][34] However, ideally, one should design a p-n homojunction, where the n-and p-type layers are composed of the same base QD materialas realized before with PbS QDs. 35 Therefore, precise control over the electrical properties of III-V semiconductor QDs, with the ability to produce robust p-type doping is crucial for the future adoption of III-V semiconductors in all-colloidal QDs based optoelectronic devices and the fabrication of InAs QDs-based p-n homojunction to compete with Pb and Hg containing counterparts. Moreover, the study and development of doped III-V colloidal QDs is interesting also from the fundamental aspects of doping in highly confined structure. As such, it is necessary to study and understand the mechanism, structural, and physical properties of the doped system.
We reported previously on an approach to dope pre-synthesized InAs QDs with metal impurities which brings them to the heavy doping regime. [19][20][21] An advantage of such postsynthesis doping reaction is the ability to separate the control of the QD size and its related initial optoelectronic characteristics, from the doping step, thus emulating the merits of the powerful and widely utilized doping processes in bulk semiconductors. This also enables to directly compare the doped versus pristine nanocrystals made from the same batch.
The effect of the post-synthesis doping approach on the performance of colloidal InAs QDsbased field effect transistors (FETs) established the ability to manipulate the doping level and the majority carrier type. InAs QDs that were post-synthetically doped with Cu presented enhanced n-type doping properties in QDs-based FETs compared to as-synthesized, undoped QDs, 22 in line with previous spectroscopic and theoretical studies. 19,20 Recently, we also developed a modified post-synthetic doping procedure of InAs QDs with Cd, achieving robust p-type doping of InAs QDs in FETs. 23 Using X-ray absorption fine structure (XAFS) spectroscopy measurements, the site-specific location of Cd within the QDs host lattice was identified, showing that it acts as a substitutional dopant to Indium near the surface of the QDs.
We also found that surface Cd not only acts as a p-dopant, but also protects the QDs against oxidation, allowing for stable device operation when exposed to air. However, while Cd-doped InAs serves as an excellent model system to study p-type doping effects in colloidal InAs QDs -implementation of such doped QDs is hindered because of the restrictions on the commercial use of Cd as defined in the RoHS.
With this in mind, we sought to develop a doping procedure using a less toxic and RoHS compliant p-type dopant for InAs QDs. Zn is known to induce p-type doping in III-V semiconductor nanowires, 36-39 thus making it a promising candidate. The incorporation of Zn during the synthesis of colloidal InP QDs has been extensively studied in terms of its positive effect on the PL quantum yield. [40][41][42][43] Reports are mixed with regards to the Zn location upon such reaction, between surface deposition to some incorporation inside the QD lattice forming an In(Zn)P alloy composition. 43 Use of less reactive Zn-carboxylate precursor was found to mostly lead to surface deposition, with some incorporation of Zn into the lattice while using more reactive shorter chain carboxylate. Recently, ZnCl2 mediated synthesis of small InAs QDs utilizing aminoarsine precursor was reported, yielding improved size distribution and enhanced fluorescence at 860 nm. 27 In this case, the Zn was deposited on the surface without incorporation into the lattice. While the possible formation of impurity states related to Zn in In(Zn)P NCs was studied theoretically, 44

Synthesis and Doping Reaction of InAs QDs
Colloidal InAs QDs were synthesized utilizing the continuous injection of InAs nanoclusters method with slight modifiactions. 28 This method enables the synthesis of high quality and large scale batches of InAs QDs with tunable band gap between 800 -1400 nm and offers delicate control over the size distribution of the as-synthesized QDs with narrow absorption and emission features. Briefly, InAs NCs seeds with excitonic absorption peak at 750 nm were synthesized by hot injection reaction between In(Oleate)3 and tris(trimethylsylil)arsine at 300c.
Then, pre-synthesized InAs nanoclusters were injected at a constant rate into the reaction flask containing the seeds for continuous, controllable growth of highly monodisperse InAs NCs (further details regarding the synthesis can be found in the supporting information). Figure 1a shows STEM characterization of a monodisperse sample of highly crystalline, as-synthesized 5 nm InAs QDs with trigonal-like shape and a narrow excitonic absorption feature at 1100 nm, indicative of the narrow size distribution.
Post synthesis doping of the as-synthesized InAs QDs with Zn was inspired by our previous work on Cd-doped InAs, 23  ). This already serves as a first indication for the substitution of Indium with Zinc.
Doping with Zn(Oleate)2 on the other hand, did not produce such byproducts.
The absorption and PL spectra of the as-synthesized InAs QDs and the purified Zn-reacted QDs are presented in figure 1c,d. In both Zn-reaction routes, the absorption feature is redshifted and broadens. However, this broadening is more pronounced for the QDs doped with diethylzinc, the more reactive precursor, implying for successful doping and higher Zn content within the QDs lattice. In the limit of heavy doping, both in bulk semiconductors as well as in nanocrystals, band tailing occurs due to the distortion in the lattice. This leads to red shifted band gap and to broadening as observed previously for heavily doped InAs NCs. 19,22,23 In addition, the PL of the diethylzinc doped QDs increases significantly, by more than an order of magnitude relative to the as-synthesized QDs and a 2-fold increase in the PL quantum yield

Field Effect Transistors Characterization
InAs QDs-based FETs were fabricated using our established method reported previously. 22  its magnitude increases as more negative gate bias is applied,  To support this hypothesis, we aimed to gain more insight on the chemistry of the doping process and characterized the as-synthesized and the doped QDs using x-ray photoelectron spectroscopy (XPS), to identify chemical changes on each of the above elements after doping.
XPS is a surface sensitive characterization method and is suitable for this study, where the main   Therefore, to confirm this hypothesis and to reveal the site-specific location of Zn within the InAs lattice, thus providing insight on the doping mechanism, we turn to atomic level structural characterization utilizing synchrotron-based XAFS measurements of the doped samples.

XAFS study for the analysis of the doping mechanism
XAFS measurements provide a powerful tool to study both the local and electronic structures of impurity doped colloidal nanocrystals. 20  concentration increases, an additional feature develops at 2 Å, which is likely to be associated with a Zn-As interaction due to substitutional In-to-Zn doping. Importantly, samples that share this additional peak also contain the highest amount of Zn, according to ICP-MS. In contrast, the Zn K-edge XANES and EXAFS data for the samples prepared using Zn(oleate)2 precursor paint a completely different picture. Figure 5d shows the Zn K-edge XANES spectra of the We support our qualitative interpretation of both doping models by quantitative fitting analyses of the EXAFS data. As informed by our visual examination of the raw data, for the sample doped using diethylzinc precursor, we used a model describing a combination of Zn as a substitutional dopant for In in the bulk of the QDs and of surface Zn forming ZnO, described by contributions from both Zn-O and Zn-As scattering paths. We hypothesize that given that the as-synthesized InAs QDs surface is Indium rich (as indicated by ICP-MS), Zn substitutes In atoms in the bulk of the QDs and will thus interact with the nearest neighboring As atoms.
In this scenario, remaining Zn cations that are exposed to the surface of the QDs and will oxidize in the presents of small amounts of oxygen and the high temperature during the reaction.
It is evidenced that for samples that contain a low concentration of Zn, using only the Zn-O path is sufficient to fit the data presented in figure 5c. However, as the Zn content within the samples increases, the Zn-As path must be considered as well to fit the data, and this path becomes increasingly more significant for higher Zn concentrations. For the Zn (oleate)2 precursor-doped samples, to fit the first coordination shell of the EXAFS data, we used a single Zn-O scattering path only (figure 5f). The model fits the low and moderate dope samples data well. For the highly doped sample, an additional Zn-As scattering path was included, however, its contribution did not affect the overall fit significantly, compared to the fit of the diethylzinc precursor-doped material. The overall picture emergesfor the diethylzinc doped samples, the higher the Zn concentration is, the more substitutional doping occurs, as depicted in figure 6a, plotting the Zn-O and Zn-As coordination numbers acquired from the fit results and illustrated in the inset. For the Zn(Oleate)2 doped InAs samples, the existence of Zn-As interaction is only evidenced in the highly doped sample with an only minor contribution, as presented in figure   6b.

The above discussed multi-technique analysis demonstrated that, upon the doping of InAs
QDs with diethylzinc, Zn diffuses into the core of the QDs, leading to substitutional doping and p-type conduction in FETs. In contrast, in Zn(Oleate)2doped QDs, Zn predominantly attaches to the surface of the QDs while also replacing some of the In on the InOx phase. Zn also resides on the surface upon the diethylzinc treatment as observed from STEM, XPS and XAFS results. This already significantly lowers the "doping efficiency", namely the extent of free carriers actually induced relative to the Zn amount. The surface Zn is not contributing free carriers. Additionally, the original high n-type character of the pristine NCs must be overcompensated to turn over to p-type conduction. This limits the doping efficiency as well.
There may also be additional factors in such small NCs, which decrease doping efficiency. This may include formation of defects, leading even to band tailing in the heavy doping limit. 19 The doping efficiency as expressed in the FET channels was also very low also in Cu doped n-type further studies of this aspect.

Conclusion
In summary, we developed a post-synthesis doping reaction to achieve heavy metal-free p-type

Supporting Information
Supporting Information is available from the Wiley Online Library or from the author.    In/In-K As/As-K Zn/Zn-K Figure S3. In K-edge normalized XANES spectra and Fourier transform magnitude of In k 2 -weighted EXAFS data of Zn(Oleate) 2 doped InAs samples    Figure S10. Raw data, baseline function and baseline subtraction of the XPS plots presented in figure 4 of the main manuscript.