Carbon doped hexagonal boron nitride epilayers have been grown by metal organic chemical vapor deposition. Photocurrent excitation spectroscopy has been utilized to probe the energy levels associated with carbon impurities in hexagonal boron nitride (h-BN). The observed transition peaks in photocurrent excitation spectra correspond well to the energy positions of the bandgap, substitutional donors (CB, carbon impurities occupying boron sites), and substitutional acceptors (CN, carbon impurities occupying nitrogen sites). From the observed transition peak positions, the derived energy level of CB donors in h-BN is ED ∼ 0.45 eV, which agrees well with the value deduced from the temperature dependent electrical resistivity. The present study further confirms that the room temperature bandgap of h-BN is about 6.42–6.45 eV, and the CN deep acceptors have an energy level of about 2.2–2.3 eV. The results also infer that carbon doping introduces both shallow donors (CB) and deep acceptors (CN) via self-compensation, and the energy level of carbon donors appears to be too deep to enable carbon as a viable candidate as an n-type dopant in h-BN epilayers.

As a member of the III-nitride wide bandgap semiconductor family, boron nitride (BN) is a much less matured material in comparison to GaN and AlN. Hexagonal boron nitride (h-BN) is highly promising as a deep-ultraviolet (DUV) photonic material exhibiting extremely high optical absorption and emission,1–8 as an ideal material for the exploration of two-dimensional heterojunction devices,9–11 and as a solid-state neutron detector material.12,13 Previous theoretical and experimental studies have revealed that the energy bandgap and the free exciton binding energy of this material are very large and are around 6.4–6.5 eV and 0.7–0.75 eV, respectively.1–8,14–21 Recent realization of wafer-scale semiconducting h-BN epilayers with high crystalline and optical qualities has opened up opportunities for the exploration of fundamental properties and emerging applications of this ultra-wide bandgap semiconductor. For instance, by employing photocurrent spectroscopy studies, the room temperature bandgap (Eg), binding energy of excitons (Ebx), and binding energy of acceptor-bound-excitons (Ebx) in h-BN epilayers have been directly determined from the observed transition peak positions to be Eg ∼ 6.42 eV, Ex ∼ 0.73 eV, and Ebx ∼ 0.2 eV, respectively.22 Neutron detectors fabricated from B-10 enriched h-BN epilayers have demonstrated the highest thermal neutron detection efficiency to date among solid-state neutron detectors of 51.4%.12 

More comprehensive studies on the impurity properties for h-BN produced under controlled growth conditions are still needed to provide input for approaches towards material quality improvement, to eliminate undesired defects, and to understand issues concerning possible doping approaches. Nitrogen vacancies (VN) and carbon impurities occupying nitrogen sites (CN) are known to be two of the most common defects in h-BN.23–31 Based on the photoluminescence (PL) emission spectroscopy studies performed on unintentionally doped h-BN epilayers grown under varying NH3 flow rates (hence with controlled VN concentrations)32,33 and theoretical insights,24–28 it is now understood that VN are shallow donors with an activation energy of about 0.1 eV, whereas CN are deep acceptors having two activation energies of about 2.3 and 1.1 eV.32,33 On the other hand, experimental studies on the properties of intentional doped impurities have been limited for h-BN.4,5,34–36 Prior studies seem to indicate that p-type conductivity is easier to realize in h-BN than in AlN via Mg doping, revealing the potential of h-BN to extending p-type III-nitride materials all the way up to 6 eV.4,5,34 The suitability of Si as an n-type dopant in h-BN has also been investigated,35,36 however, yielding controversial results with its activation energy, in one case of about 1.2 eV (Ref. 35) and of about 180 meV in another.36 A more recent theoretical study has indicated that substitutional carbon, carbon occupying boron site (CB), is a donor in monolayer h-BN.37 The calculated activation energy (ED) of CB donors in monolayer h-BN is about 0.94 eV.37 However, the binding energy of donors (and acceptors) is expected to increase as the dimensionality of h-BN scales from the bulk to the monolayer due to the enhanced in-plane overlap among carriers and impurities, as well as the reduced screening. In theory, ED will increase by a factor of 4 as the dimensionality of a semiconductor scales from the bulk to 2D, which implies that ED could be as low as 0.24 eV in bulk h-BN. This argument makes carbon a possible viable candidate for n-type doping in h-BN. In this work, we have conducted a preliminary investigation on intentional carbon doping by metal-organic chemical vapor deposition (MOCVD) growth and determined the energy levels of substitutional carbon impurities in h-BN epilayers via photocurrent excitation spectroscopy.

Epitaxial layers of carbon doped h-BN epilayers (h-BN:C) employed in this study were synthesized by MOCVD using triethylboron (TEB) and ammonia (NH3) as the precursors for B and N, respectively. Propane (C3H8) was used as the precursor for carbon doping. The basic layer structures of h-BN epilayers used in this study are shown in Fig. 1(a). A low temperature BN buffer layer of about 5 nm in thickness grown at 600 °C was first deposited on c-plane single crystal sapphire prior to the growth of the h-BN epilayer. The major benefits of this low temperature buffer layer include reducing the lattice mismatch between sapphire and h-BN, protecting the integrity of the sapphire surface at higher growth temperatures, and enhancing the adhesion of the subsequent epilayers. An undoped h-BN epilayer of about 30 nm in thickness was then deposited on the buffer layer at 1300 °C to serve as a template. This was then followed by the growth of a 30 nm thick h-BN:C epilayer at the same growth temperature. All layers were grown using hydrogen as a carrier gas. Similar to previous studies,38–40 high resolution x-ray photoelectron spectroscopy (XPS) measurements were employed to estimate the carbon doping concentration (NC) and NC is around 1 × 1021 cm−3 as determined from XPS tight scans after removing surface contaminants by Ar+ ion sputtering. Undoped h-BN epilayers with a thickness of about 60 nm were also grown at the same growth conditions for comparison measurements.

FIG. 1.

(a) Schematic layer structures of samples used in this study. (b) Room temperature photocurrent excitation spectra of h-BN and h-BN:C epilayers measured at a bias voltage of Vb = 100 V.

FIG. 1.

(a) Schematic layer structures of samples used in this study. (b) Room temperature photocurrent excitation spectra of h-BN and h-BN:C epilayers measured at a bias voltage of Vb = 100 V.

Close modal

For the photocurrent excitation spectroscopy measurements, photodetectors based on a metal-semiconductor-metal (MSM) architecture with micro-strip interdigital fingers were fabricated from h-BN and h-BN:C epilayers.7,22 The photolithography technique was used to pattern the interdigital fingers on the surface of h-BN and h-BN:C epilayers. Contacts consisting of bi-layers of Ti/Al (20 nm/30 nm) were deposited by e-beam evaporation for the application of bias voltages and the collection of the charge carriers (or photocurrent). The MSM detector has a device size of 0.5 mm × 0.5 mm with metal strips of 6 μm in width and spacing between the metal strips of 9 μm. A broad light source covering the wavelength range between 170 and 2100 nm [model E-99 laser-driven light source (LDLS) by Energetiq] coupled with a triple grating monochromator (Pro 2300 I of Acton by Research Corporation Spectra) was used as a variable wavelength excitation source. A source meter was used to apply voltages, and an electrometer was used to record the currents. The measurement system provides an overall spectral resolution of about 0.04 eV near the spectral region of 250 nm.

Figure 1(b) shows the room temperature photocurrent excitation spectra of h-BN:C and h-BN epilayers measured at a bias voltage of 100 V corresponding to an applied electric field of 1.1 × 105 V/cm. The results clearly show that carbon doping introduces additional spectroscopic features in the spectrum of h-BN:C compared to that of undoped h-BN. The transition peaks in a photocurrent excitation spectrum is a result of carriers making transitions from different energy levels to the conduction or the valence band. For instance, one can directly excite electrons from the valence band to the conduction band. In such a context, the transition peak corresponds to the bandgap and the photocurrent is resulted from both electron conduction and hole conduction. On the other hand, in doped semiconductors, one can directly excite electrons from neutral donors (D0) to the conduction band, in which case the transition peak corresponds to the energy level of donors and the photocurrent is contributed by electron conduction. If a large number of ionized donors (D+) are present, it is also possible to excite electrons from the valence band to be captured by D+ and leave behind an equal number of holes. In this case, the transition peak is the measure of the energy difference between the donor level and the valence band, and the photocurrent is contributed by hole conduction.

Figure 2 shows a replot of the photocurrent excitation spectrum of the h-BN:C sample shown in Fig. 1(b) with the energy positions marked for the dominant transitions. Transition peaks observed at about 4.2, 6.0, and 6.45 eV are clearly resolved in the spectrum of the h-BN:C sample, while the spectrum of the undoped h-BN sample shown in the bottom panel of Fig. 1(b) exhibits primarily one transition peak at about 6.45 eV. The 6.45 eV transition peak can be assigned to the direct band-to-band excitation in h-BN epilayers and corresponds well to the previously observed optical absorption edge at around 6.4 eV (Refs. 17 and 18) and is also consistent with the room temperature energy bandgap of 6.42 eV obtained from a previous photocurrent excitation spectroscopy measurement.22 The transition lines at about 4.2 eV and 6.0 eV are absent in undoped h-BN and are only associated with carbon doping. However, the difference between the observed bandgap energy at 6.45 eV and the transition line at 4.2 eV corresponds well to the previously measured energy level of CN deep level acceptors (carbon impurities occupy nitrogen sites) of about 2.3 eV within our spectral resolution limit. Therefore, it is highly plausible that the peak at 4.2 eV is due to the direct excitation of electrons from ionized CN acceptors (A) to the conduction band. This assignment corroborates previous PL emission spectroscopy studies which revealed that the donor-acceptor pair (DAP) transition involving nitrogen vacancy (VN) shallow donors and CN deep acceptors possesses a spectral peak positon at ∼4.1 eV.32,33 The 0.1 eV difference between the observed 4.2 eV peak in the photocurrent excitation spectrum shown in Fig. 2 and the PL emission peak of the DAP transition at 4.1 eV can be account for precisely by the activation energy of VN shallow donors, which is about 0.1 eV.32,33 Although the photocurrent excitation spectra shown in Figs. 1 and 2 did not cover beyond 500 nm, an emission line around 2.2 eV has been observed and considered as a room-temperature single photon emission source in a previous study.41 It is interesting to note that the 4.1–4.2 eV emission related to CN acceptors was also shown to exhibit single photon emission characteristics.42 

FIG. 2.

Replot of the room temperature photocurrent excitation spectrum of the h-BN:C epilayer measured at Vb = 100 V with energy peak positions marked.

FIG. 2.

Replot of the room temperature photocurrent excitation spectrum of the h-BN:C epilayer measured at Vb = 100 V with energy peak positions marked.

Close modal

The observation of the transition peak near 4.2 eV implies that a concomitant effect of substitutional carbon doping is self-compensation. In our case with a carbon concentration of about 1021 cm−3, the formation of carbon clusters is unlikely based on our previous studies performed on hexagonal boron nitride carbon alloys, h-(BN)1–x(C2)x, which suggested that h–(BN)1-x(C2)x epilayers with x <0.032 or x >0.95 are not phase separated at a growth temperature of 1300 °C (Refs. 39 and 40), in agreement with a theoretical calculation.43 Although most incorporated carbon atoms tend to form C-C pairs in h-BN due to the relatively large C-C bond energy of about 3.71 eV,44 the presence of a fraction of isolated neutral donors (D0) of carbon impurities occupying boron sites (CB) is expected. However, self-compensation induces a fraction of ionized CN acceptors (A) and an equal number of ionized donors (D+). This observation is also consistent with the fact that C is a group IV element and can substitute either B acting as a donor or N acting as an acceptor. We believe that the observed transition peak at about 6.0 eV is due to the direct excitation of electrons from the valence band to ionized CB donors (D+). In such a context, the energy difference between the band-to-band transition at 6.45 and the 6.0 eV transition provides a direct measure of the energy level of CB donors in h-BN epilayers, which is about ED ≈ 0.45 eV. This value is about a factor of 2 smaller than the calculated value of 0.94 eV in monolayer h-BN but larger than the expected theoretical limit of ∼0.24 eV in bulk h-BN. This is reasonable since other factors such as quasi-2D nature of h-BN epilayers and strain, as well as central-cell correction, would also modify the value of ED in h-BN epilayers. With the assignments of the observed dominant transition peaks in h-BN:C, we have constructed an energy diagram in Fig. 3 illustrating the energy bandgap and the energy levels of substitutional carbon impurities in h-BN epilayers.

FIG. 3.

Energy band diagram including the room temperature energy bandgap (Eg), CB donor energy level, and CN acceptor energy level in h-BN epilayers, directly obtained from the photocurrent excitation spectrum shown in Fig. 2.

FIG. 3.

Energy band diagram including the room temperature energy bandgap (Eg), CB donor energy level, and CN acceptor energy level in h-BN epilayers, directly obtained from the photocurrent excitation spectrum shown in Fig. 2.

Close modal

We have also attempted to determine the energy level of CB donors via the temperature dependent electrical resistivity. However, the measurements have to be carried out at above 600 K due to the high resistive nature of these samples. Figure 4(a) shows the temperature dependence of the resistivity (ρ), measured from 600 to 850 K. In order to determine the donor energy level, an Arrhenius plot of the measured resistivity is plotted in Fig. 4(b). The temperature dependent resistivity was fitted by the equation ρ(T) = ρ0exp(ED/KT), where ED denotes the energy level of CB donors in h-BN epilayers. The fitted ED value is about 0.45 eV, corroborating well with the ED value of about 0.45 eV determined from the photocurrent excitation spectrum shown in Figs. 2 and 3. The results shown in Fig. 4(b) further support our assignments of the transition peaks and the energy band diagram shown in Fig. 3. The present results have implications on intentional carbon doping in h-BN. Since the CN deep level acceptor concentration is directly related to the presence of nitrogen vacancies, we speculate that growing h-BN:C in a nitrogen-rich growth environment could help to reduce the self-compensation effect. However, other than self-compensation, our results suggest that carbon donor is still too deep to be a viable candidate as an n-type dopant in h-BN epilayers.

FIG. 4.

The temperature dependence of the electrical resistivity, ρ, of an h-BN:C epilayer.

FIG. 4.

The temperature dependence of the electrical resistivity, ρ, of an h-BN:C epilayer.

Close modal

In summary, photocurrent excitation spectroscopy has been utilized as an effective means to probe important parameters in h-BN epilayers. Transitions corresponding to the direct excitation of band-to-band, from the valence band to the CB donor level, and from the CN acceptor level to the conduction band have been directly observed. From the observed transition peak positions, the determined energy level of CB donors in h-BN is ∼0.45 eV, which agrees well with the value deduced from the temperature dependent resistivity. The present study further confirms that the room temperature bandgap of h-BN is somewhere in between 6.42 and 6.45 eV, and the CN deep acceptors have an energy level of about 2.2–2.3 eV. The determination of these parameters has important consequences on the basic understanding and potential applications of h-BN epilayers.

The MOCVD growth effort was supported by ARO (W911NF-16-1-0268) and monitored by Dr. Michael Gerhold. The detector fabrication and characterization efforts are supported by the DOE NNSA SSAA program (DE-NA0002927). Jiang and Lin are grateful to the AT&T Foundation for the support of Ed Whitacre and Linda Whitacre endowed chairs.

1.
Y.
Kubota
,
K.
Watanabe
,
O.
Tsuda
, and
T.
Taniguchi
,
Science
317
,
932
(
2007
).
2.
K.
Watanabe
,
T.
Taniguchi
,
T.
Niiyama
,
K.
Miya
, and
M.
Taniguchi
,
Nat. Photonics
3
,
591
(
2009
).
3.
K.
Watanabe
,
T.
Taniguchi
, and
H.
Kanda
,
Nat. Mater.
3
,
404
(
2004
).
4.
R.
Dahal
,
J.
Li
,
S.
Majety
,
B. N.
Pantha
,
X. K.
Cao
,
J. Y.
Lin
, and
H. X.
Jiang
,
Appl. Phys. Lett.
98
,
211110
(
2011
).
5.
S.
Majety
,
J.
Li
,
X. K.
Cao
,
R.
Dahal
,
B. N.
Pantha
,
J. Y.
Lin
, and
H. X.
Jiang
,
Appl. Phys. Lett.
100
,
061121
(
2012
).
6.
T.
Sugino
,
K.
Tanioka
,
S.
Kawasaki
, and
J.
Shirafuji
,
Jpn. J. Appl. Phys., Part 2
36
,
L463
(
1997
).
7.
J.
Li
,
S.
Majety
,
R.
Dahal
,
W. P.
Zhao
,
J. Y.
Lin
, and
H. X.
Jiang
,
Appl. Phys. Lett.
101
,
171112
(
2012
).
8.
B.
Huang
,
X. K.
Cao
,
H. X.
Jiang
,
J. Y.
Lin
, and
S. H.
Wei
,
Phys. Rev. B
86
,
155202
(
2012
).
9.
A. K.
Geim
and
I. V.
Grigorieva
,
Nature
499
,
419
(
2013
).
10.
N.
Alem
,
R.
Erni
,
C.
Kisielowski
,
M. D.
Rossell
,
W.
Gannett
, and
A.
Zettl
,
Phys. Rev. B
80
,
155425
(
2009
).
11.
C. R.
Dean
,
A. F.
Young
,
I.
Meric
,
C.
Lee
,
L.
Wang
,
S.
Sorgenfrei
,
K.
Watanabe
,
T.
Taniguchi
,
P.
Kim
,
K. L.
Shepard
, and
J.
Hone
,
Nat. Nanotechnol.
5
,
722
(
2010
).
12.
A.
Maity
,
T. C.
Doan
,
J.
Li
,
J. Y.
Lin
, and
H. X.
Jiang
,
Appl. Phys. Lett.
109
,
072101
(
2016
).
13.
T. C.
Doan
,
S.
Majety
,
S.
Grenadier
,
J.
Li
,
J. Y.
Lin
, and
H. X.
Jiang
,
Nucl. Instrum. Methods Phys. Res., Sec. A
748
,
84
(
2014
).
14.
B.
Arnaud
,
S.
Lebègue
,
P.
Rabiller
, and
M.
Alouani
,
Phys. Rev. Lett.
96
,
026402
(
2006
).
15.
B.
Arnaud
,
S.
Lebègue
,
P.
Rabiller
, and
M.
Alouani
,
Phys. Rev. Lett.
100
,
189702
(
2008
).
16.
L.
Wirtz
,
A.
Marini
, and
A.
Rubio
,
Phys. Rev. Lett.
96
,
126104
(
2006
).
17.
L.
Museur
,
G.
Brasse
,
A.
Pierret
,
S.
Maine
,
B.
Attal-Tretout
,
F.
Ducastelle
,
A.
Loiseau
,
J.
Barjon
,
K.
Watanabe
,
T.
Taniguchi
, and
A.
Kanaev
,
Phys. Status Solidi RRL
5
,
214
(
2011
).
18.
L.
Museur
,
E.
Feldbach
, and
A.
Kanaev
,
Phys. Rev. B
78
,
155204
(
2008
).
19.
K.
Watanabe
and
T.
Tanniguchi
,
Phys. Rev. B
79
,
193104
(
2009
).
20.
M. G.
Silly
,
P.
Jaffrennou
,
J.
Barjon
,
J. S.
Lauret
,
F.
Ducastelle
,
A.
Loiseau
,
E.
Obraztsova
,
B.
Attal-Tretout
, and
E.
Rosencher
,
Phys. Rev. B
75
,
085205
(
2007
).
21.
X. K.
Cao
,
B.
Clubine
,
J. H.
Edgar
,
J. Y.
Lin
, and
H. X.
Jiang
,
Appl. Phys. Lett.
103
,
191106
(
2013
).
22.
T. C.
Doan
,
J.
Li
,
J. Y.
Lin
, and
H. X.
Jiang
,
Appl. Phys. Lett.
109
,
122101
(
2016
).
23.
M.
Fanciulli
and
T. D.
Moustakas
,
Physica B
185
,
228
(
1993
).
24.
A.
Zunger
and
A.
Katzir
,
Phys. Rev.
11
,
2378
(
1975
).
25.
C.
Attaccalite
,
M.
Bockstedte
,
A.
Marini
,
A.
Rubio
, and
L.
Wirtz
,
Phys. Rev. B
83
,
144115
(
2011
).
26.
B.
Huang
and
H.
Lee
,
Phys. Rev. B
86
,
245406
(
2012
).
27.
V.
Wang
,
N.
Ma
,
H.
Mizuseki
, and
Y.
Kawazoe
,
Solid State Commun.
152
,
816
(
2012
).
28.
W.
Orellana
and
H.
Chacham
,
Phys. Rev. B
63
,
125205
(
2001
).
29.
T.
Taniguchi
and
K.
Watanabe
,
J. Cryst. Growth
303
,
525
(
2007
).
30.
I.
Jimenez
,
A. F.
Jankowski
,
L. J.
Terminello
,
D. G. J.
Sutherland
,
J. A.
Carlisle
,
G. L.
Doll
,
W. M.
Tong
,
D. K.
Shuh
, and
F. J.
Himpsel
,
Phys. Rev. B
55
,
12025
(
1997
).
31.
T. B.
Ngwenya
,
A. M.
Ukpong
, and
N.
Chetty
,
Phys. Rev. B
84
,
245425
(
2011
).
32.
X. Z.
Du
,
J.
Li
,
J. Y.
Lin
, and
H. X.
Jiang
,
Appl. Phys. Lett.
106
,
021110
(
2015
).
33.
X. Z.
Du
,
J.
Li
,
J. Y.
Lin
, and
H. X.
Jiang
,
Appl. Phys. Lett.
108
,
052106
(
2016
).
34.
H. X.
Jiang
and
J. Y.
Lin
,
Semicon. Sci. Technol.
29
,
084003
(
2014
).
35.
S.
Majety
,
T. C.
Doan
,
J.
Li
,
J. Y.
Lin
, and
H. X.
Jiang
,
AIP Adv.
3
,
122116
(
2013
).
36.
B.
He
,
M.
Qiu
,
M. F.
Yuen
, and
W. J.
Zhang
,
Appl. Phys. Lett.
105
,
012104
(
2014
).
37.
Y.
Fujimoto
and
S.
Saito
,
Phys. Rev. B
93
,
045402
(
2016
).
38.
M. R.
Uddin
,
T. C.
Doan
,
J.
Li
,
K. S.
Ziemer
,
J. Y.
Lin
, and
H. X.
Jiang
,
AIP Adv.
4
,
087141
(
2014
).
39.
M. R.
Uddin
,
S.
Majety
,
J.
Li
,
J. Y.
Lin
, and
H. X.
Jiang
,
J. Appl. Phys.
115
,
093509
(
2014
).
40.
M. R.
Uddin
,
J.
Li
,
J. Y.
Lin
, and
H. X.
Jiang
,
J. Appl. Phys.
117
,
215703
(
2015
).
41.
T. T.
Tran
,
C.
Elbadawi
,
D.
Totonjian
,
C. J.
Lobo
,
G.
Grosso
,
H.
Moon
,
D. R.
Englund
,
M. J.
Ford
,
I.
Aharonovich
, and
M.
Toth
,
ACS Nano
10
,
7331
(
2016
).
42.
R.
Bourrellier
,
S.
Meuret
,
A.
Tararan
,
O.
Ste´phan
,
M.
Kociak
,
L. H. G.
Tizei
, and
A.
Zobelli
,
Nano Lett.
16
,
4317
(
2016
).
43.
44.
S.
Bahandary
and
B.
Sanyal
,
Graphene-Boron Nitride Composite: A Material with Advanced Functionalities
(
InTech
,
2012
).