doping for new large area film-based diamond electronics

Accepted Manuscript Polycrystalline diamond films with tailored micro/nanostructure/ doping for new large area film-based diamond electronics

Jesus J. Alcantar-Peña, Elida de Obaldia, Pablo Tirado, Maria J. Arellano-Jimenez, Jose E. Ortega Aguilar, Jean F. Veyan, Miguel J. Yacaman, Yuriy Koudriavtsev, Orlando Auciello PII: DOI: Reference:

S0925-9635(18)30532-6 https://doi.org/10.1016/j.diamond.2018.11.028 DIAMAT 7272

To appear in:

Diamond & Related Materials

Received date: Revised date: Accepted date:

31 July 2018 29 November 2018 30 November 2018

Please cite this article as: Jesus J. Alcantar-Peña, Elida de Obaldia, Pablo Tirado, Maria J. Arellano-Jimenez, Jose E. Ortega Aguilar, Jean F. Veyan, Miguel J. Yacaman, Yuriy Koudriavtsev, Orlando Auciello , Polycrystalline diamond films with tailored micro/ nanostructure/doping for new large area film-based diamond electronics. Diamat (2018), https://doi.org/10.1016/j.diamond.2018.11.028

This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

ACCEPTED MANUSCRIPT Polycrystalline Diamond Films with Tailored Micro/Nanostructure/Doping for New Large Area Film-Based Diamond Electronics Jesus J. Alcantar-Peña1, 4, 7, Elida de Obaldia1,3, Pablo Tirado1, 4, Maria J. Arellano-Jimenez5, Jose E. Ortega Aguilar5, Jean F. Veyan1 , Miguel J. Yacaman5, Yuriy Koudriavtsev 6, Orlando Auciello1,2,* Department of Materials Science and Engineering, 2 Department of Bioengineering University of Texas at Dallas, Richardson, TX 75080, USA.

3

Facultad de Ciencia y Tecnología, Universidad Tecnológica de Panamá, Panamá, Panamá

4

Departamento de Investigación en Física, Universidad de Sonora, Rosales y Luis Encinas, Hermosillo, Sonora, 83000, México.

5

Department of Physics and Astronomy, University of Texas at San Antonio, San Antonio, TX 78249, USA

6

Electrical Engineering, CINVESTAV-IPN, D.F., México.

7

Dirección de Microtecnologías, Centro de Ingeniería y Desarrollo Industria (DICESI), Av. Playa Pie de la Cuesta No. 702, Desarrollo San Pablo. C.P. 76125, Santiago de Querétaro, Qro. MÉXICO.

NU

SC

RI

PT

1

MA

*Corresponding authors at: Department of Materials Science and Engineering, University of Texas at Dallas, Richardson, TX 75080, USA. E-mail address: [email protected] (O. Auciello)

D

Abstract

PT E

This paper describes processes developed to change two key electrical properties (electrical resistivity and carrier type) of ultrananocrystalline diamond (UNCD) to microcrystalline diamond (MCD) films. The results show that the electrical properties of the investigated

CE

polycrystalline diamond films depend on the grain size and plasma treated grain boundary networks interfaces and external films’ surfaces, in which hydrogen, fluorine or nitrogen can be

AC

incorporated to tailor electrical carriers-type to tune the electrical properties. Exploring the feasibility of modulating the resistivity of polycrystalline diamond films via tailoring of grain size, surface chemistry and nitrogen or fluorine incorporation into films’ grain boundaries and external surfaces may enable applications of these diamond films as active or heat dissipation layers on micro/nano-electronic devices. This work can open the pathway to enabling an industrial process for new diamond-based electronics, since polycrystalline diamond films can be grown with extreme uniformity on 300 mm diameter Si wafers, used in manufacturing of current Si-based micro/nano-electronic devices.

1

ACCEPTED MANUSCRIPT 1. INTRODUCTION Diamond films have attracted the interest of the research and technology-oriented communities because they exhibit a unique combination of properties [1], namely: a) high wear resistance, which resulted in commercial ultrananocrystalline diamond (UNCD)-coated mechanical pump seals and bearings with negligible wearing for long periods of time [2, 3]; b) highest hardness relative to any other film [1]; c) lowest friction coefficient for UNCD films [1], compared with microcrystalline diamond (MCD) and nanocrystalline diamond (NCD) and

PT

metal and ceramic coatings [1, 3, 4]; d) inertness to corrosive environments [1, 2]; e) excellent thermal conductivity (~1800 W/K·m) for MCD (≥1 μm grain size- close to single crystal

RI

diamond ~2100 W/K·m)) [1, 5], lower thermal conductivity, usable as thermal barrier, for

SC

NCD [5] and UNCD [1] films, due to large grain boundary networks; f) tunable electronic properties via boron doping in MCD [6, 7], NCD [7] and UNCD [7, 8] films, or nitrogen

NU

incorporation into grain boundaries of UNCD films, produced in the past via Microwave Plasma Chemical Vapor Deposition (MPCVD), which showed that N-UNCD films (as the films with N atoms in the grain boundaries were defined) exhibit semi-metallic conductivity

MA

(~ 250 cm-1) [1, 9-11], or tailored nanostructure from high resistivity pure UNCD films to low resistivity mixed UNCD/disordered graphene/graphite films grown by hot filament

D

chemical vapor deposition (HFCVD) [12]; and g) biocompatible properties, particularly

PT E

demonstrated for UNCD films [13].

The research reported in this paper was focused on investigating the fundamental processes required to tailor two key electrical properties (electrical resistivity and carrier type) of UNCD

CE

to MCD films, and obtain initial understanding of the underlying physics. Polycrystalline diamond films from UNCD (4-6 nm grain size) to NCD (10-100s nm grain size) to MCD (1-5 μm grain size) were grown and processed via insertion of nitrogen (N) atoms in grain

AC

boundaries and via Oxygen (O2) and Tetrafluoromethane (CF4)-based plasma treatment of the film surface, resulting in H atoms replacement on the film surface by O or F atoms, induced by reactive ion etching (RIE), to investigate the underlying physics related to carrier types and their effect in controlling the films’ electrical properties. The results show that: a) the electrical properties of the investigated polycrystalline diamond films depend on the grain size and tailored chemistry at grain boundary networks interfaces and external film’s surfaces, and b) the underlying physics controlling the changes is related to key processes as summarized below, namely:

2

ACCEPTED MANUSCRIPT 1) The electrical resistivity of “as grown” films increases (from mΩ to GΩ) as the grain size decreases to the NCD/UNCD nanostructure, with substantial grain boundary network incorporating hydrogen atoms satisfying sp and sp2 C atoms dangling bonds and reducing electrical conductivity. 2) The electrical resistivity of UNCD, NCD, and MCD films changes from that characteristic of the “as grown” films to that of films exposed to reactive ion etching (RIE) surface treatments with CF4-based plasma, which induces F-doping of the films’ surfaces, with F atoms replacing

PT

H and acting as electron donors.

RI

3) For UNCD films, nitrogen (N) atoms inserted in grain boundaries, satisfy dangling sp and sp2 C atoms bond, supplying electrons mobile through grain boundaries, inducing UNCD

SC

electrical conductivity as a n-type material with constant resistivity (mΩ) [1, 9-11]. 4) Finally, the p-type behavior of hydrogenated MCD films, can be changed into n-type

NU

behavior via RIE processing in a CF4-based plasma, due to F atoms replacing H atoms. Exploring the feasibility of modulating the resistivity of polycrystalline diamond films via

MA

tailoring of grain size and films’ surface chemistry or nitrogen incorporation into films’ grain boundaries may enable applications of these diamond films as active or heat dissipation layers on micro/nano-electronic devices. Another relevant issue related to the development of

D

diamond film-based electronics, is that currently single crystal diamond films has been used to

PT E

produce high power diamond-based electronic devices with promising performance, standing above the performance of SiC and GaN based devices [14-16]. However, the main limitation towards the implementation of industrial single crystalline diamond film-based electronics is

CE

the fact that single crystal diamond films can be grown only on relatively expensive single crystal diamond substrates with very small areas (around 20-40 mm diameter) or Iridium layers

AC

grown on crystalline MgO or SrTiO3 substrates which are also relatively expensive and fairly small (20-50 mm diameter) [17]. On the other hand, polycrystalline diamond films have been demonstrated that can be grown on up to 300 mm diameter Si wafers with high uniformity by hot filament chemical vapor deposition (HFCVD) [2] and on up to 200 mm diameter Si wafers, using microwave plasma chemical vapor deposition (MPCVD) [1], providing the bases, once the processing and properties are optimized close to those of single crystal diamond films, for insertion of polycrystalline diamond films into a transformational diamond-based electronics, once the technology is developed in the laboratory. In this sense, the results presented in this paper provide a pathway to polycrystalline diamond film-based micro/nano-electronic power devices. 3

ACCEPTED MANUSCRIPT 2. EXPERIMENTAL METHODS Diamond films were grown by HFCVD (Blue Wave Semiconductors system), using a mixture of gases: (a) 200 sccm H2 during 8hrs growth (MCD1 film), (b) 200 sccm H2 during 4hrs growth (MCD2 film), (c) 50 sccm H2 /50 sccm Ar (NCD film), (d) 10 sccm H2/90 sccm Ar (UNCD film) and (e) 10 sccm H2/90 sccm Ar/2.6 sccm N2 (electrically conductive N-UNCD film, produced with nitrogen incorporated in grain boundaries, satisfying sp/sp2 C-atoms dangling bonds and providing electrons for electrical conduction), with all films described

PT

above grown using fixed CH4 gas flow of 2 sccm at a total pressure of 10 Torr (1.3 kPa); the growth time for the NCD and UNCD films was 2 hrs. The filaments were heated to ~ 2200 ˚C

RI

to crack the CH4 molecules to produce the C-based species that grow the polycrystalline

SC

diamond films when impacting the substrate surface. The filament–substrate distance was kept at 2 cm for all grown films, since this filament-to-substrate distance was demonstrated to be

NU

optimum for growing UNCD films in the Blue Wave HFCVD system [12]. The substrate surface was seeded with diamond nanoparticles using the wet seeding process [1]. Details on

MA

the HFCVD growth process can be seen in Ref. 18.

For the work reported here, electrical properties were investigated for MCD, NCD, and UNCD films processed via O2 and CF4 RIE plasma treatments using a Technics Series 85-RIE

D

system at 100 mTorr/50W and 200 mTorr/100W for O2 and CF4 respectively, during 30

PT E

seconds up to 3.5 minutes. Nitrogen incorporation was carry out via exposure to 2.6 sccm of nitrogen (N2) gas flow during UNCD growth process and 50 sccm of N2 gas flow after growth, during cooling process. The resistivity of the films was measured using an Alessi four-point

CE

probe system, and studies of electrical carrier type (p or n), in the diamond films, were performed by fabricating metal-insulator-semiconductor (MIS)-type capacitors, incorporating

AC

a HfO2 dielectric layer, grown by atomic layer deposition (ALD) on the surface of the diamond films, and subsequently covered by the top electrodes produced by e-beam evaporation to grow a (Au (150 nm) /Cr (30 nm) multilayer on the HfO2 layer. For the studies of electrical properties of UNCD films, the HfO2 layer, inserted between the UNCD films and the top electrodes, was done to accumulate carriers at the HfO2/UNCD interface, to determine carrier type in the as deposited and plasma treated UNCD diamond films. UV-vis measurements were performed to measure the band gap for the UNCD films, and electrically conductive N-UNCD films, using a spectroscopy Ocean-Optics QE65 microscope. The microstructure (grain morphology and size) of all polycrystalline diamond films was characterized using scanning electron microscopy (SEM, ZEISS SUPRA-40) and high-resolution transmission electron microscopy

4

ACCEPTED MANUSCRIPT (HRTEM, JEOL 2100F), The samples for the HRTEM analysis were prepared via focused ion beam (FIB, FEI Nova 200) process. The type of carbon bonding in the films was analyzed by visible Raman spectroscopy, using a Raman Thermo Fisher Scientific System (model DXR2) spectrometer, with 532 nm wavelength laser beam, and with 1µm spatial resolution and 2 µm confocal depth resolution. The surface chemistry of the UNCD films was characterized by X-ray Photoelectron Spectroscopy (XPS), using a PHI Versa Probe II scanning XPS microscope, equipped with a concentric hemispherical analyzer, under ultrahigh vacuum

PT

conditions (4.8 × 10−8 Pa), using an Al Kα monochromatic X-ray source.

RI

3. RESULTS AND DISCUSSION

SC

Structure of Investigated Polycrystalline Diamond Films

AC

CE

PT E

D

MA

NU

Investigated polycrystalline diamond films included: i) MCD1 (3-5 µm grains), ii) MCD2 (~ 1µm grain size), iii) NCD (~ 500 nm grains), iv) UNCD (~ 4-6 nm grains) and v) N-UNCD (UNCD with N atoms inserted in grain boundaries). N-UNCD films with N atoms in grain boundaries were first produced in the past via Microwave Plasma Chemical Vapor Deposition (MPCVD), which showed that N-UNCD films exhibit semi-metallic conductivity (~250 cm) [1, 9-11, 19]. On the other hand, prior research on HFCVD growth of diamond films with Natoms incorporation focused on investigating the growth process, but no studies of electrical properties were reported [20, 21]. Figures 1 (a-e) show high resolution SEM images, revealing the structures of MCD1, MCD2, NCD, N-UNCD and UNCD films, respectively. However, the grain sizes of N–UNCD and UNCD films need HRTEM imaging to be resolved properly, as shown in Figure 5.

Figure 1. High resolution SEM images of MCD1 (a), MCD2 (b), NCD (c) N-UNCD (d), and UNCD (e) films.

X-ray Photoelectron Spectroscopy (XPS) Analysis of N-UNCD Films: Provides Valuable Information to Understand Electrical Performance XPS analysis of N-UNCD films is important because it provides valuable information on the chemical arrangement of N and C atoms in the films, which is critical for the electrical

5

ACCEPTED MANUSCRIPT performance. From this point of view, XPS analysis of N-UNCD films, described in this paper, over a long period of time, provides critical information to evaluate their potential for enabling long time performing electronic devices. The N-UNCD films, grown with the HFCVD process, were analyzed just after growth about four years ago, and analyzed again, recently, after long time exposure to the atmospheric environment, as part of the studies reported in this paper, performed to test the stability of the N-UNCD films as a functions of time, which is critical for films to be

PT

used for fabrication of devices, which are supposed to perform reliably over long periods of time. Figures 2 (a) to 2 (e) show XPS spectra of the N-UNCD films, obtained to investigate how stable the films are in time, from the chemical point of view, which may influence their electrical

RI

performance over long periods of time. Complementary XPS and Raman analysis were performed

SC

to characterize the chemical bonds characteristic of the N-UNCD films. During the first XPS analysis done four years ago, two areas of the sample were scanned with the X-ray beam to perform

NU

XPS analysis. N atoms were found in both areas. Subsequently, one area was bombarded with an Ar+ ion beam (1 keV) to clean the surface from adventitious C and O adsorbed from exposure to atmosphere, this being a normal cleaning procedure used in XPS analysis. Figures 2 (b) to (e)

MA

below show the XPS C 1s, N 1s, Ar 2p, and O 1s peaks recorded recently, four years after the N-UNCD film was first analyzed by XPS, on the area bombarded four years ago with a 1 keV Ar+

D

ion beam, for cleaning. The black curves (top curves in all XPS spectra in Figs. 2(b) to (e)) correspond to the XPS analysis of the N-UNCD film on the area bombarded with Ar+ ion beam

PT E

four years ago, for cleaning, and as introduced in the XPS system, recently, after being exposed to atmosphere for four years. The red and green curves correspond to the area of the N-UNCD film, bombarded four years ago with Ar+ ion beam, after two subsequent cycles of 30 seconds each, in

CE

the recent analysis, of bombardment with an Ar Cluster Ion Beam (Ar-CIB) (20 keV with 2500 Ar atoms cluster, yielding 8 eV single Ar bombardment of the surface, upon Ar cluster dissociation

AC

on impacting on the film’s surface), for low energy cleaning of adventitious C and O atoms adsorbed on the N-UNCD film’s surface, during the four years exposure to atmosphere. Figure 2 (c) shows a reduction in the N 1s peak intensity, due to N atoms being extracted from the surface of the film, due to Ar-CIB bombardment. The strong reduction of the O 1s peak towards disappearance in Fig. 2(e) is due to the cleaning process by Ar-CIB bombardment. The presence of the Ar 2p XPS peak, in Fig. 2 (d), is due to Ar atoms insertion into the diamond lattice, during the Ar+ ion beam bombardment performed during the first XPS experiment four years ago, which are inserted in the diamond lattice and remain very stable in the positions where they were inserted, as explained in detailed in Ref. 22. Even if the N1s peak decreases in intensity (Fig. 2 (c), the fact that the peak does not disappear, is an indication that N atoms are still present in the N-UNCD 6

ACCEPTED MANUSCRIPT film, four years after its growth, in the area that has been bombarded by the 1 keV Ar+ ion beam. The XPS analysis described here, correlates well with the electrical measurements done, which reveal that the N-UNCD films with N atoms incorporated in the grain boundary of the films appear fairly stable in time. Further work is underway to explore in more detail the stability of N atoms incorporation into N-UNCD films, since stability of N-UNCD films is critical for their use in high power electronic devices based on polycrystalline diamond films, as described in this paper.

PT

Special care was taken, during XPS analysis, to assure that the semi-metallic N-UNCD films were not being electrically charged during analysis, to avoid inappropriate identification of XPS peaks position in energy, which can happen, if proper neutralization of charging effects in XPS

RI

analysis of diamond films, were not done, as revealed in a recent publication from our group [22].

SC

In addition, the reader is directed to read Ref. 22, to see in detail the reason for the shift observed in the positions of the C 1s and Ar 2p XPS peaks in Fig. 2, when performing XPS analysis of

NU

diamond films. The paper recently published in the Carbon Journal [22], revealed a systematic misinterpretation in the literature, related to XPS analysis of diamond films. In this sense, XPS analysis of diamond films requires cleaning of the surface to eliminate the adventitious C and O

MA

atoms deposited on the surface of any material, including diamond films, from exposure to the atmospheric environment. The cleaning of the surface of diamond and any other material can

D

generally be done by bombardment of the surface under analysis with an Ar-Cluster Ion Beam (ACIB), involving a cluster containing 2500 Ar ions (20 keV, producing Ar ions of 8 eV each, upon

PT E

dissociation of the cluster impacting on the surface under analysis), or with an Ar+ ion beam of 2 to 5 KeV. Cleaning with an energetic A ion beam is necessary when the Ar-CIB does not clean completely the adventitious C and O atoms from the surface. Reports in the literature, revealing

CE

the problem related to the C 1s XPS peak splitting and shifting in binding energy, when performing XPS analysis of diamond films and single crystal diamond, was clarified by the systematic research

AC

reported in ref. 22, which showed that the shift in the energy position for the C1s peak is due to stress induced by the insertion of Ar atoms in the diamond lattice during the sputter cleaning with an Ar+ ion beam of 2-5 KeV. The Ar atoms inserted in the diamond lattice become associated to C atoms, inducing the energy shifts observed for the C1s and Ar 2p XPS peaks (See also Fig. 1 of ref. 22). The presence of Ar atoms inserted in the diamond lattice was confirmed by nanoscaleresolution Rutherford Backscattering (RBS) and HRTEM analysis [22]. Further details of the effects of Ar atoms-incorporation in diamond films and crystals, influencing their XPS analysis, can be found in Ref. 22.

7

PT

ACCEPTED MANUSCRIPT

NU

SC

RI

Figure 2. XPS spectra from analysis of N-UNCD film: a) XPS analysis on N-UNCD (i2014) after cleaning process by bombardment with an Ar+ ion beam (1 keV). The top black curves in Figures (b) to (e) show the signal intensity (cps) vs. X-ray-induced binding energy excitation of electrons for the C 1s, N 1s, Ar 2p, and O 1s XPS peaks, respectively, on the same area cleaned by Ar+ ion bombardment four years ago and as introduced in the XPS chamber recently. The red (middle) and green (bottom) curves in Figures (b) to (e) show the XPS peaks after cleaning the as grown N-UNCD surface with an Ar-CIB (20 keV) of 2500 Ar atoms with individual energy of 8 eV (20keV / 2500 atoms). It can be seen that after cleaning with an Ar-CIB the area, the XPS N peak (c-green curve) is similar to the XPS N peak observed in the XPS analysis in 2014 (see N peak in a).

MA

Measurement of Electrical Performance of N-UNCD and UNCD Films The measurements of electrical properties of all polycrystalline diamond films were done in a

D

closed chamber to maintain the films in the dark to avoid any possible photoelectric -induced generation of electrons and holes by light illumination. Figures 3 (a) and 3 (b) show the resistivity

PT E

(Ω.cm), mobility (cm2/V s) and carrier concentration (1/cm3) as a function of film’s temperature, during electrical measurements, to explore the potential behavior of N-UNCD and UNCD films grown on Si, as semiconducting materials, respectively. Comparison of Figs. 3 (a) and 3 (b) show

CE

the effect in electrical properties for the UNCD films with and without N atoms in the gran boundaries. Fig. 3 (b) shows that the carrier (electrons) concentration, resistivity and mobility of

AC

UNCD films are at lower levels than the corresponding parameters measured in the N-UNCD films (Fig. 3 (a)). The small increase of carrier (electrons) concentration and mobility, and reduction in resistivity observed in Figs. 3 (a) or 3 (b) can be interpreted as the effect of increasing the temperature of the films from ~13 ˚K to 350 ˚K (77 ˚C), which behave as high conductivity semiconductors. Figure 3 (c) shows the electrical conductivity vs. 1/T (T: temperature of the diamond film during electrical conductivity measurement), for a N-UNCD film (black curve) and a UNCD film (red curve), extracted from Figs. 3 (a) and 3 (b). The lower conductivity observed for 1/T > 50 for the UNCD film can be attributed to H atoms in the grain boundaries acting as trapping sites for electrons at the low temperatures (12-20 ˚K), during conductivity measurement.

8

ACCEPTED MANUSCRIPT For 1/T < 50, the higher temperatures (20-350 ˚K), during the conductivity measurements of the UNCD films, induces release of electrons from the trapping sites leading to higher conductivity [24]. On the other hand, for the N-UNCD films (black curve), no step change in electrical conductivity is observed at 1/T=50, which can be attributed to the dominant N atoms presence in the grain boundaries satisfying C atoms dangling bonds and eliminating trapping sites for

MA

NU

SC

RI

PT

electrons, as shown in prior work [1, 9-11, 19].

PT E

D

Figure 3. (a) and (b) Resistivity (Ω cm), mobility (cm2/V s) and carrier concentration (1/cm3) vs. film temperature, during measurements on N-UNCD and UNCD films, respectively, grown on Si; (c) Conductivity vs. 1/T for the same N-UNCD and UNCD films, respectively, used for the measurements shown in (a) and (b); (d) Optical light absorption density (OD*h vs. photon energy when exposing NUNCD and UNCD films, respectively, to light in the 200-900 nm wavelength.

CE

Characterization of Complementary Optical Performance of N-UNCD and UNCD Films Associated to Electrical Performance: Figure 3 (d) shows Optical light absorption density

AC

(OD*h vs. photon energy when exposing N-UNCD and UNCD films, respectively, to light in the 200-900 nm wavelength. The insert optical pictures correspond to the UNCD and N-UNCD films grown on fused-silica glass for optical measurements. Nitrogen incorporation into diamond partially degrades the optical properties reducing transparency [25, 26], which can explain the darkness observed in the N-UNCD film (Figure 3 (d) top right insert). The band gap (Eg) was calculated using a Tauc plot deduced from eq. (1): 𝐴

𝛼 = ℎ𝜈 (ℎ𝜈 − 𝐸𝑔 )𝑛

(1)

Where  is the coefficient of light absorption, ‘A’ is a constant dependent on the material, ‘h’ and ‘’ are the Planck constant and frequency of the radiation, respectively. The nature of the 9

ACCEPTED MANUSCRIPT transition is represented by ‘n’. For allowed direct and indirect transition; the values of ‘n’, for the UNCD and N-UNCD films, are 1/2 and 2, respectively. For UNCD films, the two slopes observed in the Tauc plot are correlated to different bandgaps, with the 5.19 eV attributed to crystalline diamond grains and the lower 3.62 eV attributed to higher conductivity graphite-type structure in the grain boundaries. The small reduction of the band gap from 3.62 for UNCD film to 3.43 for N-UNCD film (compare top and bottom curves in

PT

Figure 3 (d)) can be attributed to the incorporation of N atoms into the grain boundaries, satisfying sp C atoms dangling bonds (consisting of one hybrid electronic 2p orbital with diagonal bonds ; with the two unused 2p orbitals in the sp bonding ling perpendicular to

RI

180° apart -

SC

the hybrid 180°orbitals and perpendicular to each other); and also sp2 C atoms dangling bonds (electronic orbitals, which are directed in trigonal directions at 120° angles from each other in a - representing the primary bonding configuration in graphite or HOPG). The

NU

plane

unused 2p orbitals in sp2 and sp hybridization are free to form π bonds [27], which associate with the sp2 C-atoms bond hybridization that correspond to the D band (~1340 cm−1) and G band

MA

(~1588 cm−1) Raman peaks, respectively, revealed in the Raman spectra in Fig. 6 (a) (violet and green curves). The N atoms inserted in the N-UNCD film’s grain boundaries form electronic

D

donor levels with an activation energy of 1.7 eV, as demonstrated in prior work, which showed,

PT E

experimentally [1, 11] and theoretically [28], that nitrogen is favored by 3-5 eV for insertion into the gain boundaries.

Figure 4 shows the correlation between the experimental points corresponding to

CE

measurement of conductivity vs. temperature, during conductivity measurement, for UNCD and N-UNCD films, from results shown in Figure 3 (c), and the curve fitting using the nearest

AC

neighbor hopping (NNH) (Fig. 4 (a)) and variable range hopping (VRH) (Fig. 4 (b)) conduction models [29] (equation 2 and 3 below, respectively), in order to determine which model describes better the conduction mechanism of the UNCD and N-UNCD films: 𝐸

ln(𝜎) = ln(𝜎𝑜 ) − 𝑘𝑇𝐴

(2)

𝑇

1 4

ln(𝜎) = ln(𝜎𝑜 ) + (𝑇 ) 𝑀

(3)

Where 𝜎 is the electrical conductivity, 𝜎𝑜 is a constant, 𝐸𝐴 is the activation energy, 𝑘 is the Boltzmann constant, 𝑇 is the film temperature and 𝑇𝑀 is the critical temperature for each film.

10

PT

ACCEPTED MANUSCRIPT

SC

RI

Figure 4. Plot of experimental points from Fig. 2(c) and correlated curves calculated using the nearest neighbor hopping (NNH) model (a) and the variable range hopping (VRH) model (b), fitting the experimental measurements.

Activation energies of 5.64x10-4 eV and 1.39x10-3 eV were calculated for conductivities as a

NU

function of film temperature for N-UNCD and UNCD films, respectively, using the NNH model. These calculations showed relatively poor correlation between the experimental data and the

MA

calculated values for conductivity vs. temperature, in both cases, with a correlation coefficient of 0.77 for the N-UNCD and 0.85 for the UNCD films respectively, indicating that the NNH model

D

doesn’t accurately describe the conduction mechanism in these polycrystalline diamond films.

PT E

On the other hand, the VRH model resulted in a very accurate correlation with the experimental data for the conductivities as a function of film temperature for N-UNCD and UNCD films, with a correlation coefficient of 0.99 in both cases indicating that this model describes best the conduction mechanism of the N-UNCD and UNCD films. A critical temperature of 1.43 and 1.15

CE

K was obtained from the model for the N-UNCD and UNCD films, respectively. The hopping energy at 300 ˚K was estimated from the critical temperature for both films according to equation

AC

4, resulting in a hopping energy of around 3x10-3 eV for both films. 1 2

1

𝑇𝑀 2 ) 𝑇

𝐸𝐻𝑜𝑝𝑝𝑖𝑛𝑔 = 𝑘𝑇 (

(4)

Characterization of Structures of UNCD, and N-UNCD Films and Tailoring of Surface and Grain Boundaries’ Interface Chemistries: Relation to Type of Electrical Conduction Figures 5 (a) and 5 (b) show HRTEM images of UNCD and N-UNCD films, respectively. The image of the UNCD films reveals the grain size characteristic of UNCD (3-8 nm), with grains interlocked through sub-nanometer grain boundaries, as revealed in extensive prior work [1-4, 12] The image of the N-UNCD film reveals the presence of amorphous grain boundaries (in between grains) with inserted nitrogen atoms, confirming prior theoretical [28] and experimental [9-11]

11

ACCEPTED MANUSCRIPT

CE

PT E

D

MA

NU

SC

RI

PT

work, which demonstrated that N atoms are inserted in grain boundaries, as demonstrated by atom probe analysis of N-UNCD films with N atoms in the grain boundaries [30], satisfying C atoms dangling bonds and providing electrons for conduction. The bottom of Figure 5 (a) shows data from Fast Fourier Transform (FFT) analysis of the UNCD film, revealing the d-spacing corresponding to diamond grains grown with orientations (111), (220) and (311) of the film areas shown in the HRTEM image on top. Moreover, electron diffraction (top right insert in Figure 5 (a)) showed the diffraction spots corresponding to diamond (1.04 Å, 1.22 Å, and 2.04 Å). On the other hand, the bottom of Figure 5 (b) shows the Fast Fourier Transform (FFT) analysis of the N-UNCD film, revealing the d-spacing corresponding to diamond grains grown with orientations (111), (220) and (311) of the film areas shown in the HRTEM image on top. Electron diffraction (top right insert in Figure 5 (b)) showed the diffraction spots corresponding to diamond (1.04 Å, 1.28 Å, and 2.07Å) and a ring corresponding to d-spacing of 4.18 Å, which relates to stressed graphitic phase from the C-N bonding structure in the grain boundaries. The Van der Waals bond distance in the graphitic structure in grain boundaries is 3.4 Å [31], so, the increased “d” value observed in the diffraction pattern can be attributed to the stress in the lattice due to the N atoms incorporation in the grain boundaries.

AC

Figure 5. HRTEM images of UNCD (a) and N-UNCD (b) films, the latter with N atoms incorporated into de grain boundaries; The bottom of Figures (a) an (b) show FFT electron diffraction data obtained from HRTEM electron beam diffraction on N-UNCD film (top right diffraction inserted images in (a) and (b),

Plasma treatment is an efficient process to introduce quickly and efficiently dopant-atoms onto a diamond film surface. For example, hydrogen [32], oxygen [33], fluorine [34], and chlorine [35] terminated diamond surfaces were produced by H2 [35], O2, CF4 and Cl2–based [36] plasma etching treatment. In addition, previous work [36] revealed that bulk conduction of polycrystalline diamond films can increase by incorporation of hydrogen in the lattice, via heating in a hydrogen ambient, which resulted in a p-type conductivity of the diamond film, induced by hole carriers supplied when H atoms bond to C atoms. On the other hand, Fluorine (F)-terminated diamond

12

ACCEPTED MANUSCRIPT surfaces also show key surface properties, such as hydrophobicity [37], low friction coefficient [1] and n-type conductivity. Thus, RIE was used to modify the electrical properties of the MCD to UNCD films via modification of dangling bonds on the film surface. The electrical conductivity in polycrystalline diamond films can be controlled by hydrogen insertion in the lattice. In relation to this effect, it is relevant to notice that prior research also demonstrated that hydrogen incorporation into other semiconductors [38] affects the electrical

PT

conductivity of these materials, similarly as for diamond. Hydrogen passivation of dangling bonds in amorphous silicon [39], grain boundaries in polycrystalline silicon [40], donor and acceptor

RI

impurities in silicon [41], and shallow impurities in gallium phosphide [42], have been

SC

demonstrated.

Plasma-Based Reactive Ion Etching (RIE) CF4 Process to Tailor Electrical Conduction in UNCD films: In view of prior work described above, RIE with CF4 plasma were used to modify

NU

the conductivity of the diamond films described below. Figure 6 (a) shows Raman spectra from all polycrystalline diamond films grown in the research reported here. For the MCD1, MCD2 and

MA

NCD films, the sp3 bonding hybridization for C atoms in diamond (hybridization between the Sorbital and three P-orbitals forming 4 bonds linked by angular separation of 109.5°

)

D

appears as the diamond Raman peak signature at 1332 cm-1, which increases as diamond grain

PT E

size increases from NCD to MCD. The UNCD and N-UNCD films show the classic G-mode and D-mode signature broad peaks (Fig. 6 (a)) for the UNCD films (more details about Raman spectroscopy of diamond films can be seen in prior literature [1, 10-13, 43].

CE

Figure 6 (b) shows the resistivity of MCD2 films vs. RIE time exposure to CF4 plasma. MCD2 films treated with CF4 plasmas show 3 orders of magnitude resistivity increase in 30 seconds,

AC

reaching a steady state value. The change in resistivity is attributed to the replacement of hydrogen atoms by F atoms on the surface of the film. Figure 6 (c) shows X-ray Photoelectron Spectroscopy (XPS) analysis of MCD2 film before (NT) and after (CF4) RIE CF4 plasma treatment, which demonstrate the incorporation of the F atoms into the MCD2 film’s surface. Figure 6 (d) show the resistivity for all diamond films before and after RIE in a CF4 plasma. For as grown films, the resistivity increases (black curve) and after the CF4 plasma (red curve) treatment the resistivity decrease as the grain size decreases. This behavior is attributed to the fact that for MCD films the area covered by grain boundaries is order of magnitude smaller than for NCD and UNCD films, thus the content of hydrogen, in the latter, is orders of magnitude higher after the hydrogen insertion during growth process, due to hydrogen atoms low energy binding to sp and sp2 C 13

ACCEPTED MANUSCRIPT dangling bonds in the grain boundaries, as opposed to practically nil binding to saturated diamond’s sp3 C atoms bonds. The binding of hydrogen atoms to the C sp and sp2 chemical bonds is facilitated by carbon atoms forming nonpolar covalent bonds with hydrogen because carbon and hydrogen have similar electronegativities, sharing their bonding electrons almost equally. However, not all carbon atoms have the same electronegativity, that is, an sp hybridized carbon bond exhibits more electronegativity than an sp2 hybridized carbon bond, which in turn is more

PT

electronegative than an sp3 hybridized carbon bond, which is just slightly more electronegative than a hydrogen atom. Thus, after RIE in a CF4 plasma, the resistivity is strongly reduced in the

RI

UNCD films because the hydrogen atoms are replaced in the large network of grain boundaries by the F atoms binding to the sp and sp2 C bonds, providing electrons for electrical conduction

SC

through the grain boundaries. On the other hand, the resistivity of N-UNCD films is not affected by the CF4 plasma, because the N atoms strongly bonded to the sp and sp2 C atoms bonds in the

NU

grain boundaries have already provided the electrons for high electrical conductivity, as

AC

CE

PT E

D

MA

demonstrated by several groups, including ours [1, 10-12, 20, 21, 28].

Figure 6. (a) Raman spectroscopy from all diamond films (MCD1, MCD2. NCD, UNCD and N-UNCD), (b) Resistivity of MCD2 film after RIE with a CF4 plasma at different RIE times, (c) XPS analysis of MCD2 film before (NT) and after (CF4) RIE in CF4 plasma, demonstrating the incorporation of F atoms in the films, and (d) Resistivity from all diamond films before and after being exposed to a CF4 plasma.

Measurements of Electrical Performance of MCD and UNCD-Based Metal-InsulatorSemiconductor Capacitors

14

ACCEPTED MANUSCRIPT Figure 7 (a) shows pictures of MIS capacitors fabricated using MCD1 and UNCD films grown on Si wafers, to investigate how the electrical carrier content is affected after RIE in a CF4 plasma. Figure 7 (b) shows a schematic of the capacitors based on high conductivity silicon (0.001-0.005 Ω.cm) wafers bottom electrode/MCD1 and UNCD films/HfO2 dielectric layer/Cr (30 nm)/Au

SC

RI

PT

(150 nm) top electrode heterostructures.

NU

Figure 7. (a) picture of arrays of MIM capacitors fabricated with MCD and UNCD films grown on Si, using HfO2 films as dielectric layer between the top Au/Cr heterostructured electrode and the highconductivity Si-based bottom electrode; (b) Schematic of capacitors fabricated with MCD and UNCD films.

Figures 8 (a) and 8 (c) show Capacitance – Voltage (C-V) curves for UNCD and MCD films-

MA

based MIM capacitors (inserted schematics represents the capacitor device cross section), respectively, before the CF4 plasma treatment, showing a C-V curve characteristic of a p-type behavior that correlate with information about p-type hydrogenated diamond [44]. R&D on

D

Hydrogen-terminated diamond has been performed to explore developing a new technology for

PT E

p-doping of diamond, usually with a surface carrier density of 1013 cm-2 to form a two-dimensional hole gas (2DEG), that is used as p-channel on single crystal diamond field-effect-transistors (FETs), which exhibit excellent DC and RF characteristics in metal-insulator-semiconductor

CE

FETs (MISFETs) and metal-semiconductor FETs (MESFETs) [44]. The differences on the carrier modulation performance between UNCD (Fig. 8 (a)) and MCD (Fig. 8 c) films, can be attributed

AC

to hydrogen atoms (H) chemically attached to C atoms dangling bonds localized in the grain boundaries, and providing carriers for electrical current.

15

RI

PT

ACCEPTED MANUSCRIPT

NU

SC

Figure 8. Capacitance vs voltage from (a) UNCD film with no treatment, (b) UNCD film after RIE in a CF4 plasma, (c) MCD film with no treatment and (d) MCD film after RIE in a CF4 plasma. Inserted schematics represent the capacitor device cross section.

A model recently published in the literature [44] indicate that carbon-hydrogen bond (C-H

MA

bond) is essential for forming a hole accumulation layer on the surface of diamond films (see Fig. 10 (a)). The model suggests that electrons from hydrogen atoms moves to C atoms where they (electrons) bind strongly, resulting in spontaneous polarization (Fig. 10 (a)) due to the formation

D

of C-H dipoles that occurs on the hydrogen-terminated surface, owing to the difference in

PT E

electronegativity. In relation to this model, the hypothesis proposed here is that the C-H dipole formation effects on diamond surfaces, as explained above, may be happening at the surface of polycrystalline diamond grain boundaries (a detail description of this model can be found in Ref.

CE

44). UNCD films with large grain boundaries area provide enhanced carrier (positive) motion trough grain boundaries, leading to high leakage current, thus reducing charge accumulation correlated to small capacitance modulation as function of voltage, as observed in Fig. 8 (a). On

AC

the other hand, MCD films with much smaller grain boundaries area than UNCD, would exhibit lower leakage current resulting in higher charge accumulation, thus high capacitance modulation as a function of applied voltage, as observed in Fig. 8 (c). In addition to the discussion presented above, another effect observed in the capacitance modulation of the MCD film (Fig. 8 (c)), is the appearance of small flat capacitance change in the range -1 to -2 V. A potential explanation for this effect is that at -1 V there is a partial neutralization between positive charged holes and the negatively charge (O2 ) arising from

ambient humidity [44] that stabilize the accumulation of positive charged holes. Then, at -2 V,

16

ACCEPTED MANUSCRIPT the increased electric field enhance the attraction of dipper located positive holes that can contribute to increase the capacitance up to the applied voltage of -6 V. A potential explanation that accounts for the effect of negatively charge (O2 ) from humidity being strongly reduced in

UNCD films is due to the large network of grain boundaries containing large amounts of H atoms (Fig. 9(a)), from the growth process, which can cancel the chemical effect of O2 from the

atmospheric environment, since it is has been demonstrated that a strong chemical reaction

PT

between H and O atoms is produced when air interacts with H terminated diamond surfaces [45] resulting in formation of water layers [45]. In the case of Hydrogenated UNCD, it has recently

RI

been shown that the UNCD films exhibit high hydrophobicity [37], which eliminates water from

SC

the film, thus the oxygen atoms attached to the hydrogen atoms in the large hydrophobic surface grain boundary network being released. This hypothesis is strongly supported by Secondary Ion

NU

Mass Spectroscopy (SIMS) analysis of as grown UNCD and MCD films, exposed to atmospheric environment, revealing order of magnitude relative Oxygen/Hydrogen larger differential ratio

PT E

D

MA

than for MCD, as revealed by comparison of Fig. 9(a) and (b).

CE

Figure 9. SIMS analysis of as grown UNCD (a) and MCD (b) films exposed to atmospheric environment.

A relevant effect, observed from the electrical measurements, is that UNCD (Fig. 8 (b)) and

AC

MCD (Fig. 8 (d)) films, exhibit opposite capacitance change behavior after RIE in CF4 plasma. The different behavior is attributed to hydrogen atoms being removed from the UNCD film grain boundaries and replaced by F atoms, which due to their higher electronegativity (3.98) with respect to C atoms (2.5), create a spontaneous polarization due the formation of C-F dipoles that form and electron accumulation layer on the surface of the Fluorine-terminated diamond films (see Fig. 10 (b)), thus providing high conductivity with semi-metallic behavior, creating a Metal-InsulatorMetal (MIM) structured capacitor with n-type conductivity.

17

ACCEPTED MANUSCRIPT

RI

PT

Figure 10. Schematic showing the model for (a) Hydrogen and (b) Fluorine rich diamond surface spontaneous polarization, resulting in holes and electrons accumulation in the sub-surface region of diamond films.

On the other hand, H atoms in the grain boundaries of MCD films are also replaced by F atoms

SC

during the CF4 plasma treatment, but because of the order of magnitude less grain boundary area, the hypothesis is that the sites available do not provide enough F atoms incorporation to reach a

NU

semi-metallic conduction. Instead, the F atoms may provide electrons to change the P-type nature of the as grown MCD film into n-type, revealed by the increase of capacitance in the positive

MA

voltage range, when applying the voltage between the bottom and top electrodes encapsulating the MCD film.

An effect observed on capacitance as function of voltage in Fig. 8 (a) and (d) is a hysteresis

D

behavior. This hysteresis in the C-V curves, can be due to charges trapped in the dielectric/diamond

PT E

interface region. Similar C-V hysteresis curves have been observed in charge storage measurements in diamond MIS capacitors [46, 47], and other capacitance related materials like pGaN with Al2O3 dielectric layers, which were attributed also to charge trapping at the dielectric/p-

CE

GaN interface [48].

AC

4. CONCLUSIONS

The experimental results shown above indicate that the resistivity of polycrystalline diamond films depend on the grain size, N atoms incorporation in grain boundaries during growth, and RIE treatments in CF4 plasma. The data suggest that the resistivity increase in as grown polycrystalline diamond films is driven mainly by H atoms incorporation in grain boundaries, which increases substantially for NCD and UNCD films because large grain boundary networks. In relation to the effect of CF4 RIE plasma treatment, the data show that 30 seconds treatment is enough to replace H atoms by F atoms at external film surface and grain boundary interfaces, such that F atoms bonding to the C atoms release electrons to provide n-type conductivity.

18

ACCEPTED MANUSCRIPT It appears, from the data, that the grain size/grain boundaries ratio for MCD films with 3-5 µm grain size provides a very appropriate pathway to achieve carrier modulation with semiconductor behavior, being able to change the conductivity from p-type, for as grown polycrystalline diamond films, to n-type either after RIE CF4 plasma treatment or nitrogen incorporation, during film grow, into the grain boundaries. The possibility of changing the conductivity type from P to N for MCD films may open the

PT

pathway to enable new generation of polycrystalline diamond films-based electronic devices alternatively to single crystal diamond films-based counterpart. The idea of developing

RI

polycrystalline diamond films for diamond electronics is because it has been demonstrated that polycrystalline diamond films can be grown with excellent thickness uniformity and structure on

SC

up to 300 mm diameter silicon wafers, which are the bases for low cost micro/nano electronic industry, as oppose to single crystal diamond films that can be grown only on small area diamond

NU

substrates. ACKNOWLEDGMENTS

AC

CE

PT E

D

MA

Prof. O. Auciello acknowledges support from the University of Texas-Dallas, through his Distinguished Endowed Chair Professor grant. Prof. Elida de Obaldia acknowledges support from the Universidad Tecnológica de Panamá. Prof. M.J. Yacaman acknowledges support from the Welch Foundation grant AX-1615.

19

ACCEPTED MANUSCRIPT REFERENCES 1. O. Auciello and A.V. Sumant, Status review of the science and technology of ultrananocrystalline diamond (UNCD™) films and application to multifunctional devices, Diam. Relat. Mater. 19(7) (2010) 699-718. 2. Advanced Diamond Technologies (www.thindiamond.com). 3. A. V. Sumant, A. R. Krauss, D. M. Gruen, O. Auciello, A. Erdemir, M. Williams, A. F. Artiles, and W. Adams, Ultrananocrystalline diamond film as a wear-resistant and protective coating for mechanical seal applications, Tribology Trans. 48 (2005) 24-31.

RI

PT

4. R. Rani, K.J. Sankaran, K. Panda, N. Kumar, K. Ganesan, S. Chakravarty and I-Nan Lin, Tribofilm formation in ultrananocrystalline diamond film, Diam. Relat. Mater. 78 (2017) 1223.

SC

5. W. L. Liu, M. Shamsa, I. Calizo, A.A. Balandin, V. Ralchenko, A. Popovich and A. Saveliev, Thermal conduction in nanocrystalline diamond films: Effects of the grain boundary scattering and nitrogen doping, Appl. Phys. Lett. 89 (2006) 171915.

NU

6. M. Kapilashrami, G. Conti, I. Zegkinoglou, S. Nemsák, C.S. Conlon, T. Tórndahi, V. Fjállstróm, J. Lischner, S. G. Louie, R. J. Hamers, L. Zhang, J.-H. Guo, C. S. Fadley and F. J. Himpsel, Boron Doped diamond films as electron donors in photovoltaics: An X-ray absorption and hard X-ray photoemission study, J. Appl. Phys. 116 (2014) 143702.

MA

7. R. Locher, J. Wagner, F. Fuchs, C. Wild, P. Hiesinger, P. Gonon, and P. Koidl, Boron doped diamond films: Electrical and optical characterization and the effect of compensating nitrogen, Materials Science and Engineering B29 (1995) 211-215.

PT E

D

8. S. Wang, V. M. Swope, J. E. Butler, T. Feygelson, and G. M. Swain, The structural and electrochemical properties of boron-doped nanocrystalline diamond thin-film electrodes grown from Ar-rich and H2-rich source gases, Diam. Relat. Mater. 18 (2009) 669-677. 9. D.M. Gruen, A.R. Krauss, O. Auciello, and J.A. Carlisle, N-Type doping of NCD dilms with nitrogen and electrodes made therefrom, US patent #6,793,849 B1 (2004).

CE

10. J. Birrell, J. E. Gerbi, O. Auciello, J. M. Gibson, D.M. Gruen, and J.A. Carlisle, Bonding structure in nitrogen doped ultrananocrystalline diamond, J. Appl. Phys. 93 (2003) 56065612.

AC

11. S. Bhattacharyya, O. Auciello, J. Birrell, J.A. Carlisle, L.A. Curtiss, A.N. Goyette, D.M. Gruen, A.R. Krauss, J. Schlueter, A. Sumant, and P. Zapol, Synthesis and characterization of highly-conducting nitrogen-doped ultrananocrystalline diamond films, Appl. Phys. Lett. 79(10) (2001) 1441-1443. 12. J. J. Alcantar-Peña, J. Montes, M. J. Arellano-Jimenez, D. Berman-Mendoza, R. García, M.J. Yacaman and O. Auciello, Low temperature hot filament chemical vapor deposition of Ultrananocrystalline Diamond films with tunable sheet resistance for electronic power devices, Diam. Relat. Mater. 69 (2016) 207-213. 13. O. Auciello, P. Gurman, M. B. Guglielmotti, D.G. Olmedo, A. Berra and M. J. Saravia, Biocompatible ultrananocrystalline diamond coatings for implantable medical devices, MRS Bulletin, 39 (07) (2014) 621-629. 14. S. Shikata, Single crystal diamond wafers for high power electronics. Diam. Relat. Mater. 65 (2016) 168-175.

20

ACCEPTED MANUSCRIPT 15. J. W. Liu, M. Y. Liao, M. Imura, E. Watanabe, H. Oosato, and Y. Koide, Diamond logic inverter with enhancement-mode metal-insulator-semiconductor field effect transistor. Appl. Phys. Lett. 105(8) (2014) 082110. 16. M. Kasu, K. Ueda, Y. Yamauchi, A. Tallaire, T. Makimoto, Diamond-based RF power transistors: Fundamentals and applications. Diam. Relat. Mater. 16(4) (2007) 1010-1015. 17. S. T. Lee and Y. Lifshitz, Materials science: The road to diamond wafers, Nature 424 (2003) 500-5001.

RI

PT

18. E.M.A. Fuentes-Fernandez, J.J. Alcantar-Peña, G. Lee, A. Boulom, H. Phan, B. Smith, T. Nguyen, S. Sahoo, F. Ruiz-Zepeda, M.J. Arellano-Jimenez, P. Gurman, C.A. MartinezPerez, M.J. Yacaman, R.S. Katiyar, and O. Auciello, Synthesis and characterization of microcrystalline diamond to ultrananocrystalline diamond films via Hot Filament Chemical Vapor Deposition for scaling to large area applications, Thin Solid Films 603 (2016) 62-68.

SC

19. Y. Tzeng, S. Yeh, W.C. Fang, and Y, Chu, Nitrogen-incorporated ultrananocrystalline diamond and multi-layer-graphene-like hybrid carbon films, Scientific Reports, 4 (2014) 4531-4537.

NU

20. A. Afzal, C.A. Rego, W. Ahmed, and R.I. Cherry, HFCVD diamond grown with added nitrogen: film characterization and gas-phase composition studies, Diam. Relat. Mater. 7(7) (1998) 1033-1038.

MA

21. Z. Yiming, F. Larsson, and K. Larsson, Effect of CVD diamond growth by doping with nitrogen, Theoretical Chemistry Accounts, 133(2) (2013) 1432-1444.

PT E

D

22. J-F. Veyan, E de Obaldia, J. J. Alcantar-Peña, J. Montes-Gutierrez, M. J. Arellano-Jimenez, M. J. Yacaman, and O. Auciello, ,Argon atoms insertion in diamond: New insights in the identification of carbon C1s peak in X-ray photoelectron spectroscopy analysis, Carbon, 134 (2018) 29-36. 23. P. Chambrion, T. Suzuki, Z. Zhang, T. Kyotani and A. Tomita, XPS of Nitrogen-containing functional groups formed during the C−NO reaction, Energy and Fuels, 11 (3) (1997) 681685.

CE

24. K. Okano, S. Koizumi, S.R.P. Silva and G.A.J. Amaratungas, "Low-threshold cold cathodes made of nitrogen-doped chemical-vapour-deposited diamond", Nature, 381(6578) (1996) 140.

AC

25. M. E. Newton and J. M. Baker, "Models for the di-nitrogen centers found in brown diamond ", Journal of Physics: Condensed Matter, 3(20) (1991) 3605. 26. R. Samlenski, C. Haug, R. Brenn, C. Wild, R. Lecher, and P. Koidl, "Characterization and lattice location of nitrogen and boron in homoepitaxial CVD diamond", Diamond and Related Materials, 5 (1996) 947. 27. S.O. Kasap, Principles of Electronic Materials and Devices (McGraw-Hill, New York, 2002), 2nd ed. 28. J. Robertson and C.A. Davis, "Nitrogen doping of tetrahedral amorphous carbon", Diamond and Related Materials, 4(4) (1995) 441. 29. J.C. Gonzalez, G.M. Ribeiro, E. R. Viana, P.A. Fernandes, P.M.P. Salomé, K. Gutiérrez, A.F. da Cunha, " Hopping conduction and persistent photoconductivity in Cu2ZnSnS4 thin

films", Journal of Physics D: Applied Physics, 46 (2013) 155107. 21

ACCEPTED MANUSCRIPT

30. P. R. Heck, F. J. Staderman, D. Isheim, O. Auciello, T. L. Daulton, A. M. Davis, J. W. Elam, C. Floss, J Hiller, D. J. Larson, J. B. Lewis, A. Mane, M. J. Pellin, M R. Savina, D. N. Seidman, and T. Stephan, "Atom-probe analyses of nanodiamonds from Allende", Meteoritics & Planetary Science, 49(3) (2014) 453. 31. R. Andrews, D. Jacques, D. Qian and E.C. Dickey, "Purification and structural annealing of multiwalled carbon nanotubes at graphitization temperatures", Carbon, 39(11) (2001) 1681.

PT

32. P.K. Baumann and R.J. Nemanich, "Surface cleaning, electronic states and electron affinity of diamond (100), (111) and (110) surfaces", Surface Science, 409(2) (1998) 320.

RI

33. D.A. Tryk, K. Tsunozaki, T. N. Rao and A. Fujishima, "Relationships between surface character and electrochemical processes on diamond electrodes: dual roles of surface termination and near-surface hydrogen", Diamond and Related Materials, 10(9) (2001) 1804.

SC

34. K.J. Rietwyk, S. L. Wong, L. Cao, K. M. O'Donnell, L. Ley, A. T. S. Wee and C. I. Pakes, "Work function and electron affinity of the fluorine-terminated (100) diamond surface", Applied Physics Letters, 102(9) (2013) 091604.

NU

35. M. I. Landstrass and K.V. Ravi, "Hydrogen passivation of electrically active defects in diamond", Applied Physics Letters, 55(14) (1989) 1391.

MA

36. T. Kondo, H. Ito, K. Kusakabe, K. Ohkawa, K. Honda, Y. Einaga, A. Fujishima and T. Kawai, "Characterization and electrochemical properties of CF4 plasma-treated borondoped diamond surfaces", Diamond and Related Materials, 17(1) (2008) 48.

PT E

D

37. A. Gabriela Montano-Figueroa, J. J. Alcantar-Peña, P. Tirado, A. Abraham, E. de Obaldia, and O. Auciello, “Tailoring of polycrystalline diamond surfaces from hydrophilic to superhydrophobic via synergistic chemical plus micro-structuring processes”, Carbon, 139 (2018) 361-368 38. S.J. Pearton, J.W. Corbett, and T.S. Shi, "Hydrogen in crystalline semiconductors", Applied Physics A, 43(3) (1987) 153.

CE

39. K. Chandra P and R. K.V, Editors Contents of Silicon Processing in Photovoltaics I, in Materials Processing: Theory and Practices, Elsevier. p. xii (1987).

AC

40. J.I. Hanoka, C. H. Seager, D. J. Sharp and J. K. G. Panitz, "Hydrogen passivation of defects in silicon ribbon grown by the edge‐ defined film‐ fed growth process", Appl. Phys. Lett., 42(7), (1983) 618. 41. A.J. Tavendale and S.J. Pearton, "Deep level, quenched-in defects in silicon doped with gold, silver, iron, copper or nickel", Journal of Physics C: Solid State Physics, 16(9) (1983) 1665. 42. M. Singh and J. Weber, "Shallow impurity neutralization in GaP by atomic hydrogen", Appl. Phys. Lett., 54(5) (1989) 424. 43. J.J. Alcantar-Peña, G. Lee, E. M. A. Fuentes, P. Gurman, M. Quevedo, S. Sahoo, R. S. Katiyar, D. Berman and O. Auciello, "Science and technology of diamond films grown on HfO2 interface layer for transformational technologies", Diamond and Related Materials, 69 (2016) 221. 44. K. Hirama, H. Takayanagi, S. Yamauchi, Y. Jingu, H. Umezawa and H. Kawarada. Highperformance p-channel diamond MOSFETs with alumina gate insulator. in 2007 IEEE International Electron Devices Meeting. 2007.

22

ACCEPTED MANUSCRIPT

AC

CE

PT E

D

MA

NU

SC

RI

PT

45. S. Torrengo, PhD thesis, Surface functionalization and characterization of diamond thin films for sensing applications. University of Trento, 2010 p. 22-28. 46. M. Marchywka, S.C. Binaxi and D. Moses, "Observation of charge storage in diamond MIS capacitors", Electronic letters 3 (1994) 365-366. 47. M. W. Geis, J. A. Gregory, and B. B. Pate, "Capacitance-Voltage Measurements on Metal- SiO2-Diamond Structures Fabricated with (100)- and (1 1 1)-Oriented Substrates", IEEE transactions on electron devices, 38 (3) (1991) 619-626. 48. L. Sang, B. Ren, M. Liao, Y. Koide, and M. Sumiya, "Suppression in the electrical hysteresis by using CaF2 dielectric layer for p-GaN MIS capacitors", J. Appl. Phys. 123, (2018) 161423.

23

ACCEPTED MANUSCRIPT

AC

CE

PT E

D

MA

NU

SC

RI

PT

Graphical Abstract

24

ACCEPTED MANUSCRIPT Highlights 

Tailored synthesis of polycrystalline diamond films with controlled gain sizes on large area Si substrates.



Tailored dopants insertion into the films grain boundaries and surface, controlling carriers-type for tailored electrical conductivity for high performance polycrystalline

Demonstration of growth of high quality polycrystalline diamond films, with

RI

competitive electronic properties, on up to 300 mm diameter substrates to enable low

CE

PT E

D

MA

NU

SC

cost industrial process for a diamond film-based high power electronics.

AC



PT

diamond film-based electronics.

25