Valence and conduction band offsets at amorphous hexagonal boron nitride interfaces
with silicon network dielectrics
Sean W. King, Michelle M. Paquette, Joseph W. Otto, A. N. Caruso, Justin Brockman, Jeff Bielefeld, Marc
French, Markus Kuhn, and Benjamin French
Citation: Applied Physics Letters 104, 102901 (2014); doi: 10.1063/1.4867890
View online: http://dx.doi.org/10.1063/1.4867890
View Table of Contents: http://scitation.aip.org/content/aip/journal/apl/104/10?ver=pdfcov
Published by the AIP Publishing
Articles you may be interested in
Surface band bending and band alignment of plasma enhanced atomic layer deposited dielectrics on Ga- and Nface gallium nitride
J. Appl. Phys. 116, 123702 (2014); 10.1063/1.4895985
Valence and conduction band offsets at low-k a-SiOxCy:H/a-SiCxNy:H interfaces
J. Appl. Phys. 116, 113703 (2014); 10.1063/1.4895135
Valence and conduction band alignment at ScN interfaces with 3C-SiC (111) and 2H-GaN (0001)
Appl. Phys. Lett. 105, 081606 (2014); 10.1063/1.4894010
Valence band offset at the amorphous hydrogenated boron nitride-silicon (100) interface
Appl. Phys. Lett. 101, 042903 (2012); 10.1063/1.4739474
X-ray photoelectron spectroscopic evaluation of valence band offsets for strained Si 1x Ge x , Si 1y C y , and Si
1xy Ge x C y on Si(001)
Appl. Phys. Lett. 70, 2702 (1997); 10.1063/1.118998
This article is copyrighted as indicated in the article. Reuse of AIP content is subject to the terms at: http://scitation.aip.org/termsconditions. Downloaded to IP:
134.193.128.109 On: Mon, 22 Sep 2014 21:09:50
APPLIED PHYSICS LETTERS 104, 102901 (2014)
Valence and conduction band offsets at amorphous hexagonal boron
nitride interfaces with silicon network dielectrics
Sean W. King,1,a) Michelle M. Paquette,2 Joseph W. Otto,2 A. N. Caruso,2
Justin Brockman,1 Jeff Bielefeld,1 Marc French,1 Markus Kuhn,1 and Benjamin French3
1
Logic Technology Development, Intel Corporation, Hillsboro, Oregon 97124, USA
Department of Physics and Astronomy, University of Missouri-Kansas City, Kansas City, Missouri 64110, USA
3
Ocotillo Materials Laboratory, Intel Corporation, Chandler, Arizona 85248, USA
2
(Received 18 December 2013; accepted 24 February 2014; published online 11 March 2014)
To facilitate the design of heterostructure devices employing hexagonal/sp2 boron nitride, x-ray
photoelectron spectroscopy has been utilized in conjunction with prior reflection electron energy
loss spectroscopy measurements to determine the valence and conduction band offsets (VBOs and
CBOs) present at interfaces formed between amorphous hydrogenated sp2 boron nitride (a-BN:H)
and various low- and high-dielectric-constant (k) amorphous hydrogenated silicon network
dielectric materials (a-SiX:H, X ¼ O, N, C). For a-BN:H interfaces formed with wide-band-gap
a-SiO2 and low-k a-SiOC:H materials (Eg ffi 8.28.8 eV), a type I band alignment was observed
where the a-BN:H band gap (Eg ¼ 5.5 6 0.2 eV) was bracketed by a relatively large VBO and CBO
of 1.9 and 1.2 eV, respectively. Similarly, a type I alignment was observed between a-BN:H and
high-k a-SiC:H where the a-SiC:H band gap (Eg ¼ 2.6 6 0.2 eV) was bracketed by a-BN:H with
VBO and CBO of 1.0 6 0.1 and 1.9 6 0.2 eV, respectively. The addition of O or N to a-SiC:H was
observed to decrease the VBO and increase the CBO with a-BN:H. For high-k a-SiN:H
(Eg ¼ 3.3 6 0.2 eV) interfaces with a-BN:H, a slightly staggered type II band alignment was
observed with VBO and CBO of 0.1 6 0.1 and 2.3 6 0.2 eV, respectively. The measured a-BN:H
VBOs were found to be consistent with those deduced via application of the commutative and
transitive rules to VBOs reported for a-BN:H, a-SiC:H, a-SiN:H, and a-SiO2 interfaces with Si
C 2014 AIP Publishing LLC. [http://dx.doi.org/10.1063/1.4867890]
(100). V
Hexagonal/sp2 boron nitride (h-BN) is currently a material of significant interest for applications in a variety of
graphene-related devices1–3 due to excellent material
properties4 and close lattice matching with graphene (2%).5
It is additionally of considerable interest as a low-dielectricconstant (low-k) dielectric barrier material,4,6 and has
recently been demonstrated as both a gate dielectric and
effective p-type buffer layer for III–V nitride devices.7,8 In
all of these applications, h-BN forms interfaces with a variety of materials where charge transport across that interface
is either required or a significant reliability concern. In both
situations, the valence and conduction band alignment
between h-BN and the interfacing material play a role in
determining the success of the device.9,10 However, the band
alignment at many of the h-BN interfaces in the above devices has gone largely unexplored.6,11 Therefore, we have utilized x-ray photoelectron spectroscopy (XPS) to investigate
the valence band offset (VBO)/discontinuity present at the
interface between amorphous hexagonal BN and a variety of
amorphous hydrogenated silicon network dielectric materials
(a-SiX:H, X ¼ O, N, C).12,13 The conduction band offsets
(CBOs) at the BN interfaces were subsequently deduced
using the measured VBO and the band gap of the various
materials previously determined using reflection electron
energy loss spectroscopy (REELS).6,14–16
Hydrogenated amorphous BN (a-BN:H) was selected
as a proxy for h-BN in this study due to the previously
a)
Author to whom correspondence should be addressed. Electronic mail:
sean.king@intel.com. Tel.: 503-613-7547. Fax: 971-214-7811.
0003-6951/2014/104(10)/102901/4/$30.00
demonstrated sp2/hexagonal structure exhibited by this material and the ease of depositing it on a variety of different substrates and surfaces.11 The silicon network materials selected
for investigation comprised various insulating dielectric
materials that might come in contact with h-BN in the
above-mentioned devices.2 Specifically, the Si network materials included a-SiO2 and low-k a-SiOC:H materials typically
utilized as a substrate, gate oxide, or interlayer dielectric,12,17
and comparatively higher-dielectric-constant (high-k)
a-SiN:H, a-SiC:H, a-SiCO:H, and a-SiCN:H materials typically utilized as an etch stop, spacer, or diffusion barrier in
nanoelectronic devices.18 The previously reported properties
for the a-BN:H and silicon network dielectric materials investigated in this study are summarized in Table I.6,14–18
The a-BN:H thin films utilized in this study were deposited by plasma-enhanced chemical vapor deposition
(PECVD) at temperatures on the order of 400 C using various mixtures of argon, hydrogen, nitrogen, ammonia, and
diborane.6,11 In previous reports, it has been demonstrated
that these films have near 100% sp2 BN bonding, analogous to hexagonal boron nitride, with approximately 20%
hydrogen present as terminal BH and NH groups.11 The
low- and high-k dielectric silicon network materials were
similarly deposited by PECVD using a variety of different
silane, organosilane, and alkoxysilane precursors diluted in
various mixtures of hydrogen, carbon dioxide, nitrogen, and
ammonia as previously described.14–18
To help minimize charging effects such as those
observed in XPS VBO measurements for high-k dielectric/Si
interfaces,19 the dielectric/a-BN:H interfaces were prepared
104, 102901-1
C 2014 AIP Publishing LLC
V
This article is copyrighted as indicated in the article. Reuse of AIP content is subject to the terms at: http://scitation.aip.org/termsconditions. Downloaded to IP:
134.193.128.109 On: Mon, 22 Sep 2014 21:09:50
102901-2
King et al.
Appl. Phys. Lett. 104, 102901 (2014)
TABLE I. Summary of a-BN:H and low-/high-k silicon network dielectric thin-film properties.6,14–16
Film
a-BN0.7:H
a-SiN:H
a-SiC0.6N0.5:H
a-SiC:H
a-SiC0.7O0.5:H
a-SiO1.2C0.3:H
a-SiO2
Mass density (g/cm3)
Refractive index (60.001)
Young’s modulus (GPa)
Hardness (GPa)
Dielectric constant, k
Eg (eV)
1.7 6 0.1
2.6 6 0.1
2.3 6 0.1
2.5 6 0.1
2.0 6 0.1
1.5 6 0.1
2.3 6 0.1
1.73
1.99
2.03
2.55
1.77
1.44
1.46
63 6 7
185 6 18
125 6 12
210 6 20
110 6 11
19 6 2
95 6 9
861
20 6 2
17 6 2
26 6 3
14 6 2
3.0 6 0.5
10 6 1
4.2 6 0.1
6.5 6 0.1
5.8 6 0.1
7.2 6 0.1
4.8 6 0.1
3.3 6 0.1
4.2 6 0.1
5.5 6 0.2
3.3 6 0.2
3.2 6 0.2
2.6 6 0.2
3.1 6 0.2
8.2 6 0.2
8.8 6 0.2
on Cu thin films on 300 mm diameter Si (100) substrates.6
We have previously demonstrated that charging of the
a-BN:H and low-/high-k dielectrics investigated in this study
is greatly diminished when deposited on Cu substrates at the
thicknesses employed.14 Also as both classes of dielectrics
are insulators, we expect that if any small charging effects
were present in the XPS measurements, differential charging
between the two dielectrics should be below the energy resolution of the experiment (<0.1 eV). The Cu films used in this
study were formed by the physical vapor deposition of a
TaN adhesion layer and a Cu seed layer followed by Cu electrochemical plating and then a brief chemical mechanical
polish to remove surface roughness and leave a nominally
350 nm thick film.6,14
After deposition of 25 nm of a-BN:H on the Cu surface,
the wafers were transferred ex-situ to another PECVD tool
where the various low- and high-k dielectrics were deposited
at thicknesses of 5–10 nm. In all cases, an in-situ H2 plasma
treatment was performed prior to the low-/high-k dielectric
deposition in order to reduce any surface oxides or organics
formed on the a-BN:H surface during air transfer.20 Film
thickness post deposition was determined using a J.A.
Woollam variable angle spectroscopic ellipsometer.21
For x-ray photoemission studies, the dielectric/
a-BN:H/Cu samples were transferred ex-situ to a Kratos axis
HS photoemission system equipped with a hemispherical
analyzer and a monochromated Al anode x-ray source
(1486.6 eV). Surface sputtering of the samples was completed
using a 500 eV Arþ ion sputtering beam.14 The relative elemental composition/stoichiometry (not including hydrogen)
displayed in Tables I and II was obtained through analysis of
XPS survey scans using Scofield relative sensitivity factors.22,23 Core level (CL) spectra were fit with 70/30
GaussianLorentzian peaks and a Shirley24 background, and
the absolute binding energy was corrected by referencing to the
Ar 2p3/2 core level at 241.3 eV with an accuracy of 60.1 eV.
The method of Kraut et al.,25 previously described in
detail,26,27 was utilized to determine the VBO at the dielectric/
a-BN:H interface. The method relies on referencing distinct
CLs in the boron nitride and Si network dielectric materials (BN
and DL, respectively) to their respective valence band maxima
(VBM) and then measuring the relative position of these core
levels with respect to one another at their interface, as per
DEv ðDL=BNÞ ¼ ðCL VBMÞBN ðCL VBMÞDL þ DCLint ;
(1)
where DEv is the valence band offset between the two materials, (CL VBM) is the relative position of the core level to
the valence band maximum of the bulk material, and DCLint
is the relative position of the core levels in the two materials
at the interface [i.e., DCLint ¼ (CLDL CLBN)int]. To determine DCLint, we deposited 5–10 nm of low- or high-k dielectric on the a-BN:H film and measured the relative position
of the B 1s and Si 2p core levels at the interface.
(CL VBM)bulk is typically measured using thicker films
(>20 nm) to minimize or eliminate contributions to the photoemission spectra from the underlying substrate/interface.
In our case, we have previously determined (CL VBM)bulk
for a-BN:H and the low-/high-k dielectrics using the B 1s
and Si 2p core levels, respectively, detected from 25 nm
thick films separately deposited on Cu (see Table II).11,14,15
Toward establishing a qualitative estimate of VBOs, we
compare in Figures 1 and 2 XPS valence band spectra
acquired from 25 nm low-/high-k Si network and a-BN:H
films deposited on Cu. Figure 1 focuses on the nitrogencontaining materials, and as can be seen, the valence band
maxima for all three materials occur at roughly the same
position. This is consistent with the valence band of Si3N4
and BN being determined by mostly N 3p states and the conduction band being predominantly determined by Si 3p28–30
and B 2p31,32 states. Figure 2 focuses on the oxygen- and
carbon-containing materials. As one can see, the valence
TABLE II. Summary of a-BN:H and low-/high-k SiX:H (X ¼ O, C, N) XPS results and calculated a-BN:H/SiX:H valence and conduction band offsets. Note:
Negative VBO (CBO) indicates a-BN:H VBM (CBM) is below (above) the SiX:H dielectric VBM (CBM).
Film
a-BN0.7:H
a-SiN:H
a-SiC0.6N0.5:H
a-SiC:H
a-SiC0.7O0.5:H
a-SiO1.2C0.3:H
a-SiO2
B 1s VBM (eV)
Si 2p VBM (eV)
DCLint: B 1s Si 2p (eV)
VBO (eV)
CBO (eV)
189.3 6 0.1
NA
NA
NA
NA
NA
NA
NA
100.8 6 0.1
100.3 6 0.1
99.8 6 0.1
100.8 6 0.1
99.8 6 0.1
98.9 6 0.1
NA
88.4 6 0.1
89.3 6 0.1
90.5 6 0.1
88.6 6 0.1
87.7 6 0.1
88.5 6 0.1
NA
0.1 6 0.1
0.3 6 0.1
1.0 6 0.1
0.1 6 0.1
1.8 6 0.1
1.9 6 0.1
NA
2.3 6 0.2
2.0 6 0.2
1.9 6 0.2
2.3 6 0.2
0.9 6 0.2
1.4 6 0.2
This article is copyrighted as indicated in the article. Reuse of AIP content is subject to the terms at: http://scitation.aip.org/termsconditions. Downloaded to IP:
134.193.128.109 On: Mon, 22 Sep 2014 21:09:50
102901-3
King et al.
FIG. 1. XPS valence band spectra for plasma-deposited a-SiN:H, a-SiCN:H,
and a-BN:H.
band position steadily decreases in binding energy from 5 to
1 eV as the carbon content in the film increases and one goes
from pure a-SiO2 to pure a-SiC:H. This is a result of the
VBM transitioning from being predominantly determined by
O 2p states as in SiO233 to C 3p states as in SiC.34 Based on
Figures 1 and 2, one would qualitatively expect relatively
small VBOs at BN/nitride interfaces and more significant
offsets at BN/oxide and carbide interfaces.
The details of the CL–VBM measurements for a-BN:H
and the low-/high-k dielectrics have been previously
described and are summarized in Table II.6,14,15 Briefly, the
valence band maximum for each material was located by a
linear extrapolation of the leading edge of the VB emission
to the x-axis, and the energy position of the B 1s or Si 2p
core level was determined as described earlier.6,14,15 To
determine DCLint, the relative position of the Si 2p and B 1s
core levels at the a-SiX:H (X ¼ O, C, N)/BN:H interfaces
was determined after sputtering away 1–5 nm of the top
dielectric film such that any atmospheric surface contamination was removed and the B 1s core level of the buried
a-BN:H interface could be detected with adequate signal in
the XPS spectra. The Si 2p and B 1s core levels for all interfaces were well fitted using a single GaussianLorentzian
peak and a Shirley background from which the energy positions were extracted. The full width at half maximum for the
B 1s and Si 2p core level peaks were 2.0 6 0.3 eV for all
interfaces, consistent with prior measurements. This indicates a lack of significant interfacial intermixing as previously observed for a-BN:H interfaces with Si (100).11
Table II summarizes the B 1s – Si 2p DCLint values determined for each interface as well as the valence band offsets
FIG. 2. XPS valence band spectra for plasma-deposited a-SiC:H, a-SiCO:H,
a-SiOC:H, and a-SiO2.
Appl. Phys. Lett. 104, 102901 (2014)
computed using Eq. (1). As anticipated, a range of values
were observed with large offsets of 1.9 6 0.1 and 1.8 6 0.1 eV
being determined for the a-SiO2/BN:H and a-SiOC:H/BN:H
interfaces, respectively, and a relatively small offset of
0.1 6 0.1 eV being determined for the a-SiN:H/BN:H interface. An intermediate VBO of 1.0 6 0.1 eV was determined
for the a-SiC:H/BN:H interface, where the negative value
indicates that the a-BN:H VBM lies below that of a-SiC:H.
However, the addition of N or O to a-SiC:H resulted in
the VBO with a-BN:H decreasing to 0.3 6 0.1 and
0.1 6 0.1 eV, respectively, for the a-SiCN:H and a-SiCO:H
interfaces. These results are all consistent with the qualitative
alignment anticipated from Figures 1 and 2.
The VBO of 1.0 6 0.1 eV for the a-SiC:H/BN:H
interface is in reasonable agreement with a value of
0.7 6 0.2 eV recently reported by Majety et al. based on
analysis of IV measurements performed on a h-BN/6H-SiC
(0001) interface.35 Aside from this recent measurement, the
authors are unaware of prior measurements for the VBO
between BN and similar or related SiO2, Si3N4, and SiC materials. However, there have been several prior reports covering
the VBO at interfaces between these materials and Si
(001).36–45 Combining these VBO values with a prior measurement of the VBO at the a-BN:H/Si interface,11 one can use the
rules of transitivity and commutativity to deduce the VBO at
a-BN:H interfaces with SiO2, Si3N4, and SiC.46 The transitivity
and commutativity rules for VBOs, respectively, state that
DEv ð1=2Þ þ DEv ð2=3Þ þ DEv ð3=1Þ ¼ 0;
(2)
DEv ð2=3Þ ¼ DEv ð3=2Þ;
(3)
where 1/2 signifies the SiX/BN interface in question, 2/3 signifies the BN/Si interface, and 3/1 signifies the Si/SiX interface. For the a-BN:H/Si interface, we have previously
determined DEv ¼ 1.90 6 0.15 eV.11 For the SiO2/Si,
Si3N4/Si, and SiC/Si interfaces, we take DEv ¼ 4.4
6 0.2,36–39 1.6 6 0.3,39–42 and 0.7 6 0.343–45 eV, respectively, based on the average of multiple values reported in
the literature. Using these values, we determine the VBO for
BN interfaces with SiO2, Si3N4, and SiC to be 2.5, 0.3, and
1.2 6 0.3 eV, respectively. These values are in reasonable
agreement with our results and support the general trends in
VBO observed.
Having determined the valence band offsets, the CBOs
at the a-BN:H interfaces can now be determined provided
the band gaps of the two materials forming the interface are
known. The band gaps of a-BN:H and the Si network dielectrics have all been previously measured using REELS and
are summarized in Table I.6,14–16 The conduction band offsets calculated using the REELS band gaps are summarized
in Table II. The valence and conduction band offsets are also
graphically illustrated in Figure 3 for select interfaces. As
can be seen, a type I alignment with a substantial CBO is
observed for most interfaces. For the a-SiO2 and a-SiOC:H
interfaces with a-BN:H, relatively large CBOs of 1.4 6 0.2
and 0.9 6 0.2 eV were determined, respectively, where the
a-BN:H conduction band maximum (CBM) lies below the
a-SiO2 and a-SiOC:H CBM. For the a-BN:H interface with
the silicon nitrides and carbides, relatively large CBO values
of 1.9–2.3 eV were also observed with the a-BN:H CBM this
This article is copyrighted as indicated in the article. Reuse of AIP content is subject to the terms at: http://scitation.aip.org/termsconditions. Downloaded to IP:
134.193.128.109 On: Mon, 22 Sep 2014 21:09:50
102901-4
King et al.
Appl. Phys. Lett. 104, 102901 (2014)
2
K. Kim, J. Choi, T. Kim, S. Cho, and H. Chung, Nature 479, 338 (2011).
K. Watanabe, T. Taniguchi, and H. Kanda, Nature Mater. 3, 404 (2004).
S. King, H. Simka, D. Herr, H. Akinaga, and M. Garner, APL Mater. 1,
40701 (2013).
5
K. Kim, A. Hsu, X. Jia, S. Kim, Y. Shi, M. Dresselhaus, T. Palacios, and J.
Kong, ACS Nano 6, 8583 (2012).
6
S. King, M. French, J. Bielefeld, M. Jaehnig, M. Kuhn, and B. French,
Electrochem. Solid-State Lett. 14, H478 (2011).
7
S. Majety, J. Li, X. Cao, R. Dahal, B. Pantha, J. Lin, and H. Jiang, Appl.
Phys. Lett. 100, 61121 (2012).
8
T. Nguyen, H. Shi, M. Kudo, and T. Suzuki, Phys. Status Solidi C 10,
1401 (2013).
9
J. Robertson, J. Vac. Sci. Technol. A 31, 50821 (2013).
10
G. Ribes, J. Mitard, M. Denais, S. Bruyere, F. Monsieur, C. Parthasarathy,
E. Vincent, and G. Ghibaudo, IEEE Trans. Device Mater. Reliab. 5, 5
(2005).
11
S. King, M. French, J. Bielefeld, M. Jaehnig, M. Kuhn, G. Xu, and B.
French, Appl. Phys. Lett. 101, 42903 (2012).
12
K. Maex, M. Baklanov, D. Shamiryan, F. Iacopi, S. Brongersma, and Z.
Yanovitskaya, J. Appl. Phys. 93, 8793 (2003).
13
G. Wilk, R. Wallace, and J. Anthony, J. Appl. Phys. 89, 5243 (2001).
14
S. King, M. French, M. Jaehnig, M. Kuhn, B. Boyanov, and B. French,
J. Vac. Sci. Technol. B 29, 051207 (2011).
15
S. King, M. French, M. Jaehnig, M. Kuhn, and B. French, Appl. Phys.
Lett. 99, 202903 (2011).
16
S. King, B. French, and E. Mays, J. Appl. Phys. 113, 44109 (2013).
17
S. King, E. Mays, C. Ege, N. Hellgren, J. Xu, H. Li, and B. Boyanov,
Mater. Res. Soc. Symp. Proc. 1371, imrc11-1371-s1-p132 (2012).
18
S. King, D. Jacob, D. Vanleuven, B. Colvin, J. Kelly, M. French, J.
Bielefeld, D. Dutta, M. Liu, and D. Gidley, ECS J. Solid State Sci.
Technol. 1, N115 (2012).
19
E. Bersch, M. Di, S. Consiglio, R. Clark, G. Leusink, and A. Diebold,
J. Appl. Phys. 107, 43702 (2010).
20
X. Liu, S. Gill, F. Tang, S. King, and R. Nemanich, J. Vac. Sci. Technol. B
30, 31212 (2012).
21
S. King, J. Vac. Sci. Technol. A 29, 41501 (2011).
22
J. Scofield, J. Electron. Spectrosc. Relat. Phenom. 8, 129 (1976).
23
S. Hofmann, Auger- and X-Ray Photoelectron Spectroscopy in Materials
Science: A User-Oriented Guide (Springer, New York, 2013).
24
D. Shirley, Phys. Rev. B 5, 4709 (1972).
25
E. Kraut, R. Grant, J. Waldrop, and S. Kowalczyk, Phys. Rev. Lett. 44,
1620 (1980).
26
S. King, R. Davis, C. Ronning, M. Benjamin, and R. Nemanich, J. Appl.
Phys. 86, 4483 (1999).
27
S. King, R. Davis, C. Ronning, and R. Nemanich, J. Electron. Mater. 28,
L34 (1999).
28
E. Ferreira and C. Goncalves da Silva, Phys. Rev. B 32, 8332 (1985).
29
R. Karcher, L. Ley, and R. Johnson, Phys. Rev. B 30, 1896 (1984).
30
S. Ren and W. Ching, Phys. Rev. B 23, 5454 (1981).
31
Y. Xu and W. Ching, Phys. Rev. B 44, 7787 (1991).
32
J. Robertson, Phys. Rev. B 29, 2131 (1984).
33
E. Calabrese and W. Fowler, Phys. Rev. B 18, 2888 (1978).
34
J. Robertson, Philos. Mag. B 66, 615 (1992).
35
S. Majety, J. Li, W. Zhao, B. Huang, S. Wei, J. Lin, and H. Jiang, Appl.
Phys. Lett. 102, 213505 (2013).
36
E. Bersch, S. Rangan, and R. Bartynski, Phys. Rev. B 78, 85114 (2008).
37
R. Williams, J. Vac. Sci. Technol. 14, 1106 (1977).
38
F. Bell and L. Ley, Phys. Rev. B 37, 8383 (1988).
39
J. Keister, J. Rowe, J. Kolodziej, H. Niimi, T. Madey, and G. Lucovsky,
J. Vac. Sci. Technol. B 17, 1831 (1999).
40
A. Goodman, Appl. Phys. Lett. 13, 275 (1968).
41
D. DiMaria and P. Arnett, Appl. Phys. Lett. 26, 711 (1975).
42
A. Iqbal, W. Jackson, C. Tsai, J. Allen, and C. Bates, J. Appl. Phys. 61,
2947 (1987).
43
C. Bittencourt, M. De Seta, and F. Evangelisti, J. Vac. Sci. Technol. B 16,
1599 (1998).
44
R. Fang and L. Ley, Phys. Rev. B 40, 3818 (1989).
45
J. Schafer, J. Ristein, S. Miyazaki, and L. Ley, J. Vac. Sci. Technol. A 15,
408 (1997).
46
A. Katnani and R. Bauer, Phys. Rev. B 33, 1106 (1986).
47
S. Nguyen, T. Nguyen, H. Treichel, and O. Spindler, J. Electrochem. Soc.
141, 1633 (1994).
3
4
FIG. 3. Schematic of band alignment for (a) a-SiO2/a-BN:H/a-SiN:H interfaces and (b) a-SiO1.2C0.3:H/a-BN:H/a-SiC:H interfaces.
time residing above that of the SiN/SiC dielectrics. The large
CBOs are primarily due to the low VBO and relatively small
band gap (3 eV) for the SiN and SiC materials investigated.
However, it should be noted that larger band gaps of 5 eV
have been reported for silicon nitrides, where the N/Si ratios
are closer to the ideal stoichiometry of 4/3 (i.e., Si3N4).40,42
Similarly, a substantial range in band gaps of 4.7–6 eV has
been reported for h-BN materials.6,47 Thus, significant engineering of the CBO at a-SiN:H/BN:H interfaces can be
anticipated by tuning the N/Si and N/B ratios in the a-SiN:H
and a-BN:H dielectrics.
In summary, we have utilized XPS and REELS to investigate the valence and conduction band offsets present at
interfaces between a-BN:H and a variety of low- and high-k
dielectric Si network materials. The a-SiO2/BN:H interface
was found to have a type I band alignment with a large VBO
and CBO of 1.9 6 0.1 and 1.4 6 0.2 eV, respectively. In contrast, the a-SiN:H/BN:H interface was found to have a type
II alignment with a relatively small VBO of 0.1 6 0.1 eV,
but a large CBO of 2.3 6 0.2 eV. The a-SiC:H/BN:H interface was found to have a type I alignment with a VBO and
CBO of 1.0 6 0.1 and 1.9 6 0.2 eV, respectively. The intermediate VBO values observed for interfaces with
a-SiCO:H and a-SiCN:H indicate that significant tuning of
the VBO and CBO with a-BN:H can be achieved via varying
the composition of the silicon network dielectric.
1
C. Dean, A. Young, I. Meric, C. Lee, L. Wang, S. Sorgenfrei, K.
Watanabe, T. Taniguchi, P. Kim, K. Shepard, and J. Hone, Nat.
Nanotechnol. 5, 722 (2010).
This article is copyrighted as indicated in the article. Reuse of AIP content is subject to the terms at: http://scitation.aip.org/termsconditions. Downloaded to IP:
134.193.128.109 On: Mon, 22 Sep 2014 21:09:50