A thin layer of Al2O3 was employed as an interfacial layer between surface conductive hydrogen-terminated (H-terminated) diamond and MoO3 to increase the distance between the hole accumulation layer in diamond and negatively charged states in the acceptor layer and, thus, reduce the Coulomb scattering and increase the hole mobility. The valence band offsets are found to be 2.7 and 3.1 eV for Al2O3/H-terminated diamond and MoO3/H-terminated diamond, respectively. Compared to the MoO3/H-terminated diamond structure, a higher hole mobility was achieved with Al2O3 inserted as an interface layer. This work provides a strategy to achieve increased hole mobility of surface conductive diamond by using optimal interlayer along with high high electron affinity surface acceptor materials.

Diamond is a promising candidate for high frequency and high-power electronic devices due to its high thermal conductivity, wide bandgap, high carrier mobility, and high-saturated drift velocity. Field effect transistors (FETs) based on dielectric layers on hydrogen terminated diamond (H-diamond) show surface conductivity usually ascribed to the formation of a hole accumulation layer.1 The prior studies, taken together, indicate that charge transfer at the dielectric/H-diamond interface can lead to a hole accumulation layer but with relatively low mobility attributed to interface scattering from the transferred negative charge.2 A recent study of FETs from Al2O3/H-diamond has shown high RF performance with an fT of 30 GHz.1 A notable aspect of this study is the indication of a carrier velocity of 1 × 107 cm/s, which approaches the saturation velocity of diamond.3 A second set of papers focused on the low field hole mobility when hexagonal boron nitride (h-BN) is used as the dielectric on H-diamond.4–6 A high channel mobility of ∼680 cm2/V-s was achieved.6 This significant increase in the measured mobility was ascribed to effective reduction of charge scattering at the interface.5,6 Following the indication of the role of interface scattering, this report presents a different approach to increase the low field channel mobility at the dielectric/H-diamond interface following a strategy similar to modulation doping in III–V heterostructures.7 

While p- and n-type substitutional dopants exhibit high activation energies in diamond, a p-type surface conducting layer can be readily achieved by surface transfer doping at a diamond surface or interface.1 The conductivity of the surface layer is not thermally activated, which provides an advantage relative to impurity doping in diamond. Indeed, surface transfer doping provides an alternative strategy to enable high power and high frequency diamond field effect transistor (FET) operation. While surface transfer doping was first observed for air exposed H-diamond surfaces, a number of studies have shown surface transfer doping characteristics at dielectric/H-diamond interfaces.8 The surface conducting layer of both air exposed and dielectric/H-diamond interfaces can reach the following electrical properties: a sheet resistance of 5–20 kΩ/sq, a hole density of 1012–1013 cm−2, and a hole mobility of 10–150 cm2/V-s.8 

Hydrogen-terminated (H-terminated) diamond (Eg = 5.47 eV) exhibits a negative electron affinity (NEA) of −1.1 to −1.3 eV.9 Consequently, the valence band maximum (VBM) of the H-terminated diamond is ∼4.3 eV below the vacuum level. Materials with sufficiently high electron affinity (χ > 4.3 eV), such as MoO3, V2O5, WO3, could align with their conduction band minimum (CBM) below the VBM of H-terminated diamond. Electron transfer, from the diamond to the transition metal oxides, would then be energetically favorable. The relative band alignment of the dielectric layer provides the driving electric potential for the transfer of electron from the diamond valence band to the dielectric layer conduction band or localized defects within the gap.10,11

While surface conductive diamond may be enabled by high electron affinity metal oxides, a correlation is often observed between hole concentration and mobility where an increase in carrier concentration is associated with a mobility decrease or vice versa.8 One possibility is that for charge transfer doping, the density of the negatively charged states is not distributed homogenously in the oxide, but accumulate in a layer adjacent to the hole accumulation layer in the diamond.12 Those negatively charged states give rise to Coulomb scattering in the hole accumulation layer. For bulk-doping of semiconductors, the scattering of free charge carriers at low temperatures is dominated by Coulomb scattering from the ionized impurities. With the Brooks–Herring and Conwell–Weisskopf approximations, the mobility is described by13 

μI=82q3(kπ)32εr2T3/2(m*)0.5NIf,
(1)

where k is the Boltzmann constant, q is the electron elementary charge, εr is the relative permittivity, m* is the effective mass, NI is the concentration of ionized impurities, and f is the screening factor. The Coulomb scattering mechanism results in a carrier mobility that is inversely proportional to the density of charged scattering centers. Similarly, the sheet resistance is described as

Rs=1/μq(NIt),
(2)

where t is the layer thickness and (NI t) is the sheet carrier concentration. For a specific sheet resistance, μqRs = 1/(NI t) provides the relation of the mobility and sheet charge concentration for a range of values. A larger sheet charge concentration is linearly balanced by a reduction of the mobility.

For charge transfer doping, the results from temperature-dependent Hall effect measurements also confirm this.12 As the density of negatively charged states in the adsorbate or dielectric layer is approximately equal to the hole sheet concentration in the diamond, an increase in the hole sheet concentration is associated with an increase in negatively charged states in the adsorbate or dielectric layer. Furthermore, the increased density of negatively charge states would result in a reduction of hole mobility due to the Coulomb scattering.

On the other hand, the potential of a charged scattering center v(r) is ∼1/|r|. When the negatively charged states accumulate adjacent to the hole accumulation layer, the effect of Coulomb scattering is significant and limits the hole mobility.2,4 This effect has also been analyzed in detail for modulation doping in AlGaAs/GaAs heterostructures. Here, impurity doping in the AlGaAs layer can provide electrons to the GaAs charge accumulation layer, which reduces the ionized impurity scattering, and the mobility is projected to increase as d5/2, where d is the location of the dopants relative to the interface.7 Following the example of modulation doping in AlGaAs/GaAs, we propose to employ a thin layer of Al2O3 as an interfacial layer between diamond and MoO3 to increase the distance between the hole accumulation layer in diamond and the negatively charged states in the acceptor layer.

This study presents photoemission measurements of the electronic band alignment of the MoO3/Al2O3/H-diamond layer structure to gain insight into the driving potential for the negative charge transfer and the location of the negative charges near the interface, in the Al2O3 layer or in the MoO3 layer.

The diamond substrates used in this study were 5 × 5 mm2 type IIa chemical vapor deposited single crystalline undoped diamond (100) substrates. A hydrogen plasma treatment was applied for 15 min to obtain an H-terminated surface. Hydrogen terminated surfaces were prepared in a microwave plasma chemical vapor deposition system used for diamond layer growth. The process was applied for 15 min using a microwave power of 1000 W and H2 pressure of 50 Torr. The diamond substrate reached ∼800 °C due to heating from the plasma.

The Al2O3 deposition was accomplished by remote plasma enhanced atomic layer epitaxy (PEALD). The precursor used was dimethylaluminum isoproxide (DMAI, [(CH3)2AlOCH(CH3)2]2), and an O2 plasma was used as the oxidizer.14 The oxygen plasma was excited with 30 W rf power. The growth rate was ∼1.5 Å/cycle at a substrate temperature of 100 °C.

After Al2O3 deposition, the surface was processed with a hydrogen plasma post-deposition treatment for 30 min at ∼500 °C. The hydrogen plasma was excited by 100 W rf power. The chamber pressure was maintained at 100 mTorr with a constant H2 gas flow of 20 sccm. Additional details are presented in our previous work.14 

Molybdenum oxide layers were deposited in a reactive electron beam deposition (EBD) system, which has a base pressure of 1 × 10−8 Torr. Molybdenum pellets with a purity of 99.99% were evaporated using an e-beam. After deposition the substrates were transferred (in UHV) to the PEALD system where an O-plasma process was employed. The MoO3 (+6) oxidation state was achieved using a 5 min oxygen plasma treatment at ambient temperature. The remote plasma was generated by 30 W rf excitation with a 100 mTorr oxygen pressure. The oxidation state of the molybdenum oxide layers was examined with XPS.

The hole accumulation layer is associated with upward band bending in the diamond near surface region. Consequently, the surface conducting state can be indicated by measuring the binding energy of the C 1s core level. After each process (Al2O3 deposition, hydrogen plasma, Mo deposition, oxygen plasma), the surface was characterized using in situ XPS, which employed a monochromated Al Kα x-ray source (hν = 1486.6 eV) with a bandwidth of 0.2 eV and a R3000 Scienta analyzer with a resolution of 0.1 eV. All core level spectra were recorded with a 0.05 eV step, and the peak positions can be resolved to ∼±0.05 eV by curve fitting.

Ex situ Hall effect measurements were performed using an Ecopia HMS-3000 Hall measurement system. The diamond was mounted on a Van der Pauw–Hall probe. A ∼0.5 mm diameter dot of gold foil was used under each probe tip to increase the contact area. The surface resistance, carrier density, and carrier mobility were measured with a 0.51 T permanent magnet at room temperature.

Diamond samples using MoO3 as acceptor layer were prepared with 0, 2, and 4 nm Al2O3 interface layers. For the H-terminated diamond surface with air exposure, the C 1s core level was located at a characteristic position 284.1 eV, indicating the surface conductive state.14 After 2 nm MoO3 deposition by EBD, the spectrum of the Mo 3d3/2 and 3d5/2 spectral region shows three peaks, which is a superposition from metallic Mo regions and partially oxidized Mo regions [Fig. 1(c)]. The peak near 228 eV corresponds to the 3d5/2 core level of metallic Mo regions, and the central peak is a combination of the 3d3/2 from the metallic regions and the 3d5/2 of partially oxidized Mo. The third peak at ∼235 eV is the 3d3/2 of the partially oxidized Mo. After the oxygen plasma process, the molybdenum is fully oxidized, and the remining two peaks correspond to the Mo 3d3/2 and 3d5/2 core levels shifted to higher binding energies of 235.7 and 232.5 eV. These values indicate Mo in the +6 oxidation state (for MoO3) in Fig. 1 as well as the samples described in Figs. 2 and 3. The C 1s core level was not affected by the MoO3 deposition. Figure 1 shows the C 1s, O 1s, and Mo 3d core levels of the diamond surface after each process.

FIG. 1.

XPS scans of MoO3/H-diamond after each process, showing the (a) O 1s, (b) C 1s, and (c) Mo 3d core level peaks. The results indicate MoO3 after the oxygen plasma.

FIG. 1.

XPS scans of MoO3/H-diamond after each process, showing the (a) O 1s, (b) C 1s, and (c) Mo 3d core level peaks. The results indicate MoO3 after the oxygen plasma.

Close modal
FIG. 2.

XPS scans of MoO3/2 nm Al2O3/H-diamond after each process step, showing the (a) O 1s, (b) C 1s, (c) Mo 3d, and (d) Al 2p core level peaks. The results indicate MoO3 and diamond surface conductivity after the O-plasma.

FIG. 2.

XPS scans of MoO3/2 nm Al2O3/H-diamond after each process step, showing the (a) O 1s, (b) C 1s, (c) Mo 3d, and (d) Al 2p core level peaks. The results indicate MoO3 and diamond surface conductivity after the O-plasma.

Close modal
FIG. 3.

XPS scans of MoO3/4 nm Al2O3/H-diamond after each process step, showing the (a) O 1s, (b) C 1s, (c) Mo 3d, and (d) Al 2p core level peaks. The results indicate MoO3 and diamond surface conductivity after the O-plasma.

FIG. 3.

XPS scans of MoO3/4 nm Al2O3/H-diamond after each process step, showing the (a) O 1s, (b) C 1s, (c) Mo 3d, and (d) Al 2p core level peaks. The results indicate MoO3 and diamond surface conductivity after the O-plasma.

Close modal

For the diamond samples that employed 2 nm Al2O3 interfacial layers, the C 1s core level shifted to higher binding energy after Al2O3 PEALD, as shown in Fig. 2. With a hydrogen plasma post-deposition treatment, the C 1s core level was restored to 284.0 eV, the characteristic binding energy of the surface conducting state, consistent with our previous study.14 The Al 2p core level shifted 2.1 eV to a higher binding energy 75.3 eV, indicating that the hydrogen plasma process not only restored the hole accumulation layer in diamond but also resulted in a negatively charged layer at the interface. The negatively charged interfacial layer, which was adjacent to the diamond hole accumulation layer, would introduce the Coulomb scattering of the holes, limiting the hole mobility. After MoO3 deposition, the Al 2p core levels shifted 0.9 eV to lower binding energy, indicating a reduced density of negative charges close to the diamond surface. Apparently, the Coulomb scattering due to the negatively charged interfacial layer was reduced. The C 1s core level position of the diamond was not affected by the MoO3 deposition, confirming the diamond surface remained in the conductive state.

For the diamond samples with a 4 nm Al2O3 interfacial layer, after hydrogen plasma treatment, the C 1s core level at 284.2 eV (as shown in Fig. 3) was not fully restored (to 284.0 eV) characteristic of the surface conducting layer. The result may indicate that the hydrogen plasma is less effective restoring the surface conductivity for the thicker Al2O3 layers. However, with the addition of 2 nm of MoO3, the C 1s core level shifted from 284.2 to 284.0 eV, indicating electron transfer from the diamond to the MoO3 acceptor layer.

Table I summarizes the change in sheet resistance, hole concentration, and hole mobility of H-terminated diamond with MoO3 as acceptor layer and Al2O3 as an interface layer. The air-exposed, hydrogen terminated surfaces exhibit sheet resistance consistent with many other reports.1 The properties of the surface, such as roughness and hydrogen termination, contribute to the charge density and mobility of the specific surfaces.15 For the sample without Al2O3, the surface conductivity of diamond was achieved with MoO3 as an acceptor layer. After MoO3 deposition and oxidation, the sheet resistance increased to 6.9 kΩ/sq and remained under 10 kΩ/sq. The hole concentration remained the same as the H-terminated surface while the hole mobility decreased to 8.9 cm2/(V s). For the samples with an Al2O3 interface layer, the surface conductivity was achieved with the diamond/Al2O3/MoO3 structure. The sheet resistance increased similarly as without Al2O3 to 6.2 and 14.0 kΩ/sq, and the hole concentration decreased to 9.4 × 1012/cm2 and 3.5 × 1012/cm2, for the samples with 2 and 4 nm Al2O3, respectively. The hole mobility increased significantly to >100 cm2/V-s. The change in sheet resistance, hole concentration, and hole mobility for the different structures with processes are plotted in Fig. 4(a), and the hole concentration and mobility of different structures are compared in Fig. 4(b). The hole mobility and concentration vary with the change in the thickness of Al2O3.

TABLE I.

Hall measurement sheet resistance, carrier concentration, and carrier mobility of diamond surfaces using MoO3 as acceptor layer with 0, 2, and 4 nm Al2O3 interfacial layers.

SampleProcessSheet resistance (kΩ/sq)Carrier concentration/cm2Carrier mobility cm2/V-s
0 nm Al2O3 interfacial layer H-terminated surface 1.8 1.9 × 1014 18.4 
After molybdenum dep. 3.6 1.3 × 1014 13.5 
Oxidation by plasma 6.9 1.0 × 1014 8.9 
2 nm Al2O3 interfacial layer H-terminated surface 2.3 7.0 × 1014 3.94 
2 nm Al2O3 dep. and hydrogen plasma 2.0 6.4 × 1014 4.94 
Molybdenum dep. and oxidation by plasma 6.2 9.4 × 1012 100.7 
4 nm Al2O3 interfacial layer H-terminated surface 5.9 4.6 × 1013 22.9 
4 nm Al2O3 dep. and hydrogen plasma 11.5 6.7 × 1013 8.1 
Molybdenum dep. and oxidation by plasma 14.0 3.5 × 1012 125.9 
SampleProcessSheet resistance (kΩ/sq)Carrier concentration/cm2Carrier mobility cm2/V-s
0 nm Al2O3 interfacial layer H-terminated surface 1.8 1.9 × 1014 18.4 
After molybdenum dep. 3.6 1.3 × 1014 13.5 
Oxidation by plasma 6.9 1.0 × 1014 8.9 
2 nm Al2O3 interfacial layer H-terminated surface 2.3 7.0 × 1014 3.94 
2 nm Al2O3 dep. and hydrogen plasma 2.0 6.4 × 1014 4.94 
Molybdenum dep. and oxidation by plasma 6.2 9.4 × 1012 100.7 
4 nm Al2O3 interfacial layer H-terminated surface 5.9 4.6 × 1013 22.9 
4 nm Al2O3 dep. and hydrogen plasma 11.5 6.7 × 1013 8.1 
Molybdenum dep. and oxidation by plasma 14.0 3.5 × 1012 125.9 
FIG. 4.

(a) The change in sheet resistance, hole concentration, and hole mobility with processes for MoO3/0/2/4 nm Al2O3/H-diamond. (b) The hole concentration and mobility of different structures. The lines indicate constant sheet resistance.

FIG. 4.

(a) The change in sheet resistance, hole concentration, and hole mobility with processes for MoO3/0/2/4 nm Al2O3/H-diamond. (b) The hole concentration and mobility of different structures. The lines indicate constant sheet resistance.

Close modal

The valence band offset (ΔEV) between diamond and Al2O3 can be calculated using the following equation:16,17

ΔEV=(ECLEV)DiamondECLEVAl2O3ΔECL,
(3)

where ECL is the binding energy of the XPS core level; EV is the valence band maximum (VBM); (ECL − EV) is the binding energy difference from the VBM to the respective core level; and ΔECL is the binding energy difference between the diamond and Al2O3 core levels measured at the interface (ECLC1s − ECLAl2p). The value of (ECL−EV)Al2O3 was 70.6 ± 0.1 eV,18 and that of (ECL−EV)Diamond was 284.1 ± 0.1 eV. The diamond value was based on the assumption that the VBM is at the Fermi level for the surface conductive state. The valence band offset (VBO) between Al2O3 and diamond was determined to be 2.6 eV by using Eq. (3), and taking account a band bending of 1.2 eV, which resulted from a concentration of defects or interstitial oxygen atoms in Al2O3 introduced by oxygen plasma.14 The ΔEV between diamond and MoO3 was calculated to be 3.1 eV by using Eq. (3), where (ECL−EV)MoO3 was 229.5 eV as the Fermi level of MoO3 is located close to its conduction band minimum (CBM).

The band diagrams of Al2O3/H-diamond and MoO3/Al2O3/H-diamond are shown in Fig. 5. The charge at the Al2O3/H-diamond interface is represented by a decrease in the band displacement. After MoO3 was deposited on Al2O3, the Al 2p core level shifted to lower binding energy. This result indicated that the distribution of negative charges changed and part of the negative charges originally close to diamond interface transferred into the MoO3 layer.

FIG. 5.

Band diagrams of MoO3/H-diamond, Al2O3/H-diamond, and MoO3/Al2O3/H-diamond. Excess charges are indicated as well as the hole accumulation layer near the diamond interface.

FIG. 5.

Band diagrams of MoO3/H-diamond, Al2O3/H-diamond, and MoO3/Al2O3/H-diamond. Excess charges are indicated as well as the hole accumulation layer near the diamond interface.

Close modal

According to the position of the Al 2p core level, the distribution of negative charge changes with the increase in the Al2O3 layer thickness. As the charge transfer doping is driven by the energy level difference between the diamond VBM and the acceptor level, the sheet concentration and mobility of the hole accumulation layer are related to the oxide electron affinity. A higher electron affinity of the acceptor material corresponds to a larger energy difference with the diamond valence band. Evidently, oxides with a higher electron affinity lead to a higher sheet hole concentration and a lower hole mobility. Without the Al2O3 interfacial layer, the sheet concentration and mobility of the hole accumulation layer achieved with MoO3 are mainly determined by its electron affinity. With a thin Al2O3 interfacial layer, the MoO3 acts as an additional acceptor layer, where a fraction of the negative charges are located at the Al2O3/H-diamond interface and the other negative charges are distributed in the MoO3. When the Al2O3 is sufficiently thick, the negatively charged interfacial layer limits the hole mobility by Coulomb scattering.

We note that reducing the negative charge in the Al2O3 could substantially improve the efficiency of charge transfer away from the interface. Our previous study explored a 500 °C anneal after the Al2O3 ALD deposition and H-plasma process, but a reduction in the interface charge was not detected.14 Other approaches for Al2O3 deposition, such as high temperature ALD, could potentially improve the process.19 We also note that effects related to interface roughness, defects, and impurities could play a significant role in the charge transfer process.20 

For an Al2O3 interlayer, as the thickness increases from 0 to 4 nm, the hole concentration decreases and the mobility increases. This effect could be quantified with Hall measurements, which are planned for the future. The hole concentration decrease is because the distance between the diamond and acceptor layer increases, while the potential difference between the diamond and oxide bands (i.e., band offset) is unchanged. If the charge distribution is simplified to all holes accumulated at the interface of Al2O3/H-diamond and all negative charged states accumulated at the interface of Al2O3/MoO3, the hole concentration in diamond can be estimated by comparison with the ideal capacitor relations: Q = CV = (εA/d) V, where C is the capacitance, ε is the dielectric constant, A is the contact area, d is the Al2O3 thickness, and V is the (before equilibrium) energy difference between the H-diamond VBM and the MoO3 conduction band. In this case, an increase in Al2O3 thickness results in a decrease in hole concentration. Although in the MoO3/Al2O3/H-diamond structure, the negatively charged states distribute at both the Al2O3/H-diamond interface and the MoO3/Al2O3 interface. The increase in the distance between H-diamond and MoO3 results in a decrease in the charge transfer between H-diamond and MoO3.

Insight into the charge distribution can be established with capacitance–voltage measurements as has been demonstrated in recent studies of dielectric layers on H-terminated diamond.21,22 Such studies are being considered for future studies of dielectric layers on H-diamond.

The increase in the hole mobility may be attributed to the reduced Coulomb scattering with the increase in the distance between the hole accumulation layer and the negative charged states. At a certain point, the increase in the Al2O3 layer thickness limits the charge transfer to the MoO3 structure, and the interface turns into the case of thick Al2O3 on H-diamond.14 

In this work, a surface conducting layer was achieved using MoO3 as a charge transfer layer with a thin Al2O3 interfacial layer. The VBO are found to be 2.6 and 3.1 eV for Al2O3/H-terminated diamond and MoO3/H-diamond, respectively. The band alignment provides the driving potential for charge transfer. The diamond sheet resistance, hole concentration, and mobility were characterized for MoO3/H-diamond with Al2O3 interfacial layers of 0, 2, and 4 nm thickness. By combining two oxides (Al2O3 and MoO3), the hole mobility and concentration were modulated by altering the thickness of the interfacial layer. The mobility of the hole accumulation layer on diamond was improved by inserting an interfacial layer to reduce interface scattering. The overall sheet resistance for the different Al2O3 interlayers remained constant, which suggests that the charge at the Al2O3/H-diamond interface still limited the mobility. Improvement of the diamond Al2O3/H-diamond interface guided by structural and chemical analysis and Hall measurements could substantially impact the interface mobility. The results also motivate further study of alternative dielectric materials, with lower defect density, to serve as interlayers between the charge transfer dielectric and H-diamond.

This research was supported by a grant from MIT-Lincoln Laboratories and the NSF through Grant Nos. DMR-1710551 and DMR-2003567.

The authors have no conflicts to disclose.

The data that support the findings of this study are available from the corresponding author upon reasonable request.

1.
K. G.
Crawford
,
I.
Maini
,
D. A.
Macdonald
, and
D. A. J.
Moran
, “
Surface transfer doping of diamond: A review
,”
Prog. Surf. Sci.
96
,
100613
(
2021
).
2.
G.
Daligou
and
J.
Pernot
, “
2D hole gas mobility at diamond/insulator interface
,”
Appl. Phys. Lett.
116
,
162105
(
2020
).
3.
S.
Imanishi
,
K.
Horikawa
,
N.
Oi
,
S.
Okubo
,
T.
Kageura
,
A.
Hiraiwa
, and
H.
Kawarada
, “
3.8 W/mm RF power density for ALD Al2O3-based two-dimensional hole gas diamond MOSFET operating at saturation velocity
,”
IEEE Electron Device Lett.
40
,
279
(
2019
).
4.
Y.
Sasama
,
T.
Kageura
,
K.
Komatsu
,
S.
Moriyama
,
J.
Inoue
,
M.
Imura
,
K.
Watanabe
,
T.
Taniguchi
,
T.
Uchihashi
, and
Y.
Takahide
, “
Charge-carrier mobility in hydrogen-terminated diamond field-effect transistors
,”
J. Appl. Phys.
127
,
185707
(
2020
).
5.
Y.
Sasama
,
K.
Komatsu
,
S.
Moriyama
,
M.
Imura
,
T.
Teraji
,
K.
Watanabe
,
T.
Taniguchi
,
T.
Uchihashi
, and
Y.
Takahide
, “
High-mobility diamond field effect transistor with monocrystalline h-BN gate dielectric
,”
APL Mater.
6
,
111105
(
2018
).
6.
Y.
Sasama
,
T.
Kageura
,
M.
Imura
,
K.
Watanabe
,
T.
Taniguchi
,
T.
Uchihashi
, and
Y.
Takahide
, “
High-mobility p-channel wide-bandgap transistors based on hydrogen-terminated diamond/hexagonal boron nitride heterostructures
,”
Nat. Electron.
5
,
37
44
(
2022
).
7.
K.
Lee
,
M. S.
Shur
,
T. J.
Drummond
, and
H.
Morkoc
, “
Low field mobility of 2-d electron gas in modulation doped AlxGa1−xAs/GaAs layers
,”
J. Appl. Phys.
54
,
6432
(
1983
).
8.
C.
Verona
,
W.
Ciccognani
,
S.
Colangeli
,
E.
Limiti
,
M.
Marinelli
, and
G.
Verona-Rinati
, “
Comparative investigation of surface transfer doping of hydrogen terminated diamond by high electron affinity insulators
,”
J. Appl. Phys.
120
,
025104
(
2016
).
9.
D.
Takeuchi
,
H.
Kato
,
G. S.
Ri
,
T.
Yamada
,
P. R.
Vinod
,
D.
Hwang
,
C. E.
Nebel
,
H.
Okushi
, and
S.
Yamasaki
, “
Direct observation of negative electron affinity in hydrogen-terminated diamond surfaces
,”
Appl. Phys. Lett.
86
,
152103
(
2005
).
10.
S. A. O.
Russell
,
L.
Cao
,
D.
Qi
,
A.
Tallaire
,
K. G.
Crawford
,
A. T. S.
Wee
, and
D. A. J.
Moran
, “
Surface transfer doping of diamond by MoO3: A combined spectroscopic and Hall measurement study
,”
Appl. Phys. Lett.
103
,
202112
(
2013
).
11.
K. G.
Crawford
,
L.
Cao
,
D.
Qi
,
A.
Tallaire
,
E.
Limiti
,
C.
Verona
,
A.
Wee
, and
D. A. J.
Moran
, “
Enhanced surface transfer doping of diamond by V2O5 with improved thermal stability
,”
Appl. Phys. Lett.
108
,
042103
(
2016
).
12.
C. E.
Nebel
, “
Surface transfer-doping of H-terminated diamond with adsorbates
,”
New Diamond Front. Carbon Technol.
15
,
247
(
2005
).
13.
K.
Somogyi
, “
Classical approximations for ionised impurity scattering applied to diamond monocrystals
,”
Diamond Relat. Mater.
11
,
686
(
2002
).
14.
Y.
Yang
,
F. A.
Koeck
,
M.
Dutta
,
X.
Wang
,
S.
Chowdhury
, and
R. J.
Nemanich
, “
Al2O3 dielectric layers on H-terminated diamond: Controlling surface conductivity
,”
J. Appl. Phys.
122
,
155304
(
2017
).
15.
T.
Wade
,
M. W.
Geis
,
T. H.
Fedynyshyn
,
S. A.
Vitale
,
J. O.
Varghese
,
D. M.
Lennon
,
T. A.
Grotjohn
,
R. J.
Nemanich
, and
M. A.
Hollis
, “
Effect of surface roughness and H–termination chemistry on diamond's semiconducting surface conductance
,”
Diamond Relat. Mater.
76
,
79
(
2017
).
16.
J. R.
Waldrop
and
R. W.
Grant
, “
Measurement of AlN/GaN (0001) heterojunction band offsets by x-ray photoemission spectroscopy
,”
Appl. Phys. Lett.
68
,
2879
(
1996
).
17.
E. A.
Kraut
,
R. W.
Grant
,
J. R.
Waldrop
, and
S. P.
Kowalczyk
, “
Interface contributions to heterojunction band discontinuities: X-ray photoemission spectroscopy investigations
,” in
Heterojunction Band Discontinuities: Physics and Device Applications
, edited by
F.
Capasso
and
G.
Margaritondo
(
Elsevier
,
New York
,
1987
).
18.
J.
Yang
,
B. S.
Eller
,
M.
Kaur
, and
R. J.
Nemanich
, “
Characterization of plasma-enhanced atomic layer deposition of Al2O3 using dimethylaluminum isopropoxide
,”
J. Vac. Sci. Technol., A
32
,
21514
(
2014
).
19.
A.
Daicho
,
T.
Saito
,
S.
Kurihara
,
A.
Hiraiwa
, and
H.
Kawarada
, “
High-reliability passivation of hydrogen-terminated diamond surface by atomic layer deposition of Al2O3
,”
J. Appl. Phys.
115
,
223711
(
2014
).
20.
D.
Oing
,
M.
Geller
,
A.
Lorke
, and
N.
Wöhrl
, “
Tunable carrier density and high mobility of two-dimensional hole gases on diamond: The role of oxygen adsorption and surface roughness
,”
Diamond Relat. Mater.
97
,
107450
(
2019
).
21.
P. B.
Shah
,
J.
Weil
,
A. G.
Birdwell
, and
T.
Ivanov
, “
Charge trapping analysis of high speed diamond FETs
,”
MRS Adv.
2
,
2235
(
2017
).
22.
N. C.
Saha
and
M.
Kasu
, “
Heterointerface properties of diamond MOS structures studied using capacitance-voltage and conductance-frequency measurements
,”
Diamond Relat. Mater.
91
,
219
(
2019
).