DOI: 10.1002/cnma.

201900632 Communication

1
2
3 Charge Distribution and Stability of SiO2 Nanoarray Electret
4
5 Fei Liang,[a, b, c] Hua Yang Li,[a, b, c, d] Ying Wang,[a, b, c] Shuang Yang Kuang,[a, c] You Jun Fan,[a, c]
6 Zhong Lin Wang,[a, e] and Guang Zhu*[a, d]
7
8
charges in the SiO2 electret are readily dissipated in the moist
9 Abstract: The devices based on SiO2 electret have significant
environment as a result of their low volume resistivity (1016 Ω cm)
10 applications in various MEMS sensors. However, the charge
and strong hydrophilic surface.[7] So, in the last decades, extensive
11 stability of SiO2 electret suffered from the water percolation,
approaches have been reported in developing the SiO2 electrets
12 which seriously restricts its application. In this work, the long-
with long-term charge stability in harsh environments such as
13 term charge stability of the SiO2 nanoarray electret (SiO2 NAE)
hydrophobic coating,[8–9] ion implantation[10] and hydrophobic
14 without any water repellent treatment is demonstrated. When
structure.[11] Although these efforts can efficiently improve the
15 the oxidation time is 1.5 h, the potential decay of the SiO2
stability of the trapped charges after exposure to moist environ-
16 NAE is less than 46% during 60 days with an original potential
ments, the toxicity of organic coating and the complicated and
17 of 120 V. The long-term charge stability of the SiO2 NAE is
delicate fabrication process seriously constrain their real applica-
18 attributed to its unique charge decay process and the large
tions. As we all know, it is a well validated fact that the surface
19 H2O diffusion barrier in the SiO2 NAE: firstly, the charge
nanostructures of the material have a significant effect on its
20 trapped in the planar part of the SiO2 NAE decays rapidly.
surface properties.[12–14] Particularly, one-dimension nanostructure
21 Then, residual charge stored in the SiO2 nanoarray. Eventually,
(such as nanowire, nanorod and nanobelt) show a unique surface
22 the large H2O diffusion barrier in the interface of Si/SiO2
property, which can been extensively applied in optoelectronic
23 effectively hinders the charge decay. In addition, we demon-
and electronic systems.[15–19] Furthermore, these properties can
24 strate the unique charge stability of nanoelectret, which has a
highly affect the charge stability of the electrets.[20–21] Therefore,
25 promising application in developing high performance elec-
the charge stability for the electrets might be tailored by its
26 tret-based devices.
nanostructre. However, this association has yet to be applied in
27
optimizing the charge stability of electrets.
28
Herein, we develop a SiO2 nanoarray electret (SiO2 NAE)
29
An electret is a kind of insulator that exhibits a quasi-permanent with long-term charge stability even in the moist environments.
30
source of electric field. Due to the trapped electret charges, it has The SiO2 NAE is fabricated by dry oxidizing of the Si nanoarray.
31
been widely used in MEMS devices such as energy harvester,[1–2] The surface potential decay of the SiO2 NAE (oxidized for 1.5 h)
32
various transducer[3–5] and microphones.[6] Among all the electret is less than 46% after 60 days with the initial potential of
33
materials, SiO2 show the great potential in fabricating high 120 V. In addition, scanning Kevin probe microscopy (SKPM)
34
performance MEMS devices because of the high dielectric strength and finite element method (FEM) simulation are used to
35
and the compatibility in CMOS processing. However, the trapped determine the charge distribution in the SiO2 NAE. And two
36
kinds of charge distribution are observed in the SiO2 NAE after
37
corona charging, namely high potential region (HPR) and low
38 [a] F. Liang, H. Y. Li, Y. Wang, S. Y. Kuang, Y. J. Fan, Prof. Z. L. Wang,
Prof. G. Zhu potential region (LPR). Additionally, total amount of the charges
39
CAS Center for Excellence in Nanoscience trapped in LPR is 40% of the charges in HPR. During the charge
40 Beijing Key Laboratory of Micro-nano Energy and Sensor decay process, the LPR is gradual converted into the HPR. And
41 Beijing Institute of Nanoenergy and Nanosystems
Chinese Academy of Sciences then, the residual charges store on the SiO2 nanoarray. Finally,
42
Beijing 100083 (China) the large H2O diffusion barrier in the interface between Si and
43 E-mail: zhuguang@[Link] SiO2 effectively hinders the decay of charges. So, the SiO2 NAE
44 [b] F. Liang, H. Y. Li, Y. Wang
reported here could provide a new insight to develop high
45 CAS Center for Excellence in Nanoscience
National Center for Nanoscience and Technology (NCNST) performance nanoelectrets with long-term charge stability.
46
Beijing 100190 (China) The fabrication process of the SiO2 nanoarray electret (NAE)
47 [c] F. Liang, H. Y. Li, Y. Wang, S. Y. Kuang, Y. J. Fan is illustrated in Figure 1a. Firstly, the Si nanoarray was prepared
48 School of Nanoscience and Technology
University of Chinese Academy of Sciences through a metal-assisted chemically etching method.[22] The
49
Beijing 100049 (China) detail process was described in experiment section. Subse-
50
[d] H. Y. Li, Prof. G. Zhu quently, the Si nanoarray was oxidized at 1000 °C (with the O2
51 New Materials Institute, Department of Mechanical, Materials and
flow of 100 sccm) to obtain the SiO2 NAE. Finally, the SiO2 NAE
52 Manufacturing Engineering
University of Nottingham Ningbo China was charged via a corona charging process. The corresponding
53
Ningbo 315100 (China) scanning electron microscope (SEM) image of the Si nanoarray
54 [e] Prof. Z. L. Wang is present in Figure 1b, which clearly shows that the nanoarray
55 School of Materials Science and Engineering
Georgia Institute of Technology Atlanta has a uniform distribution. The mono-crystallinity of the Si
56
GA 30332 (USA) nanoarray is also demonstrated according to the high-resolu-
57
ChemNanoMat 2020, 6, 212 – 217 212 © 2019 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
 Communication

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
Figure 1. (a) Fabrication process of SiO2 NAE. (b) SEM image of Si nanoarray. (c) HRTEM image of Si nanowire, insert image is the SAED image. (d) TEM image
16
of SiO2 nanowire. (e) Corresponding EDS mapping image.
17
18
19
tion transmission electron microscope (HRTEM) image in Fig- that the same charge distribution in the SiO2 NAE with different
20
ure 1c. As shown in Figure 1d, the diameter of the Si nanowire initial potentials. And this characteristic charge distribution in
21
oxidized for 1.5 h is 260 nm, and the formation of SiO2 after the SiO2 NAE will be elaborated as bellow.
22
oxidization is also confirmed by the energy dispersive spec- To further investigate the charge distribution in the SiO2 NAE
23
trometer (EDS) elemental mapping of Si and O presented in after corona charging, the scanning Kevin probe microscopy
24
Figure 1e. (SKPM) was used to measure the surface potential distribution in
25
The oxidation time of Si nanowires may affect the surface the top surface and the cross section. Considering the low
26
potential and charge stability of the SiO2 NAE because of the measurement range of SKPM (from 10 V to + 10 V) and the
27
thickness change. First, as plotted in Figure 2 a, the surface same charge distribution in the SiO2 NAEs with different initial
28
potential of SiO2 NAE changes from 85 V to 134 V with the potentials, the sample oxidized for 1.5 h with the initial surface
29
oxidation time increasing. This result is consistent with the potential of 9 V was selected in the SKPM measurement. The
30
thickness dependence of surface potential for the SiO2 film SEM image and atomic force microscope (AFM) topography image
31
electrets.[23] Second, the surface potential decay curves in in Figure 2c and 2d demonstrate the surface morphology of the
32
Figure 2b reveal that the oxidation time has a significant effect SiO2 NAE. The different surface potentials along with the surface
33
on the charge stability of SiO2 NAE. For all the SiO2 NAEs, the height is illustrated in Figure 2e, which is confirmed by the 3D
34
trapped charges dramatically decrease within the first 8 days surface potential image (shown in Figure 2f). Moreover, as shown
35
and then remain approximately stable. Besides, it can be found in Figure 2g, it can be found that the deeper surface possesses the
36
that the optimum oxidation time is determined to be 1.5 h higher surface potential, which indicates the existence of nano-
37
because even after 60 days, the decay of the surface potential is wires on the surface has great influence on the charge distribution
38
less than 46%, which is superior to the planar SiO2 electrets in SiO2 NAE.
39
prepared by thermal oxidation (P-SiO2-Dry oxidation) and Although the surface potentials vary with the surface
40
plasma enhanced chemical vapor deposition (P-SiO2-PECVD). height, the range of the measured height in Figure 2g is less
41
Both of the two control samples have the same thickness as the than 300 nm, which is much lower than the real length of the
42
SiO2 NAE oxidized for 1.5 h (130 nm), which is confirmed by the prepared SiO2 nanowire (7 μm). Thus, to completely demon-
43
SEM images in Figure S1. Then, Figure S2a schematically strate the true charge distribution in the SiO2 NAE, the cross
44
illustrates the surface potentials of the SiO2 NAEs (oxidized for section potential distribution of the SiO2 NAE was also
45
1.5 h) with different voltages and time during the corona conducted as illustrated in Figure 3a. Meanwhile, the corre-
46
charging process. It can be found that the surface potential sponding cross-sectional SEM image and AFM topography are
47
becomes saturated ( 135 V) as the voltage and time increase. clearly presented in Figure 3b and 3c, which shows that the
48
These results are consistent with the traditional planar electrets length of the SiO2 nanowire is about 7 μm. The corresponding
49
film after the corona charging.[24–25] Subsequently, the charge surface potential distribution image in the initial stage (Fig-
50
decay curves of the SiO2 NAEs (oxidized for 1.5 h) with different ure 3d) shows that both the high potential region (HPR,
51
initial potentials are exhibited in Figure S2b. Considerable denoted as section A, C and E) and low potential region (LPR,
52
charge decay can be found as a higher initial potential in the denoted as section B and D) exist in the SiO2 NAE simulta-
53
samples, which is induced by the large charge drift rate in the neously. In the HPR, the middle of the SiO2 nanowire has a high
54
high initial potential samples.[26] Particularly, all the three surface potential while both the upper and bottom sides show
55
samples with high ( 135 V), medium ( 70 V) and low ( 9 V) a low surface potential. However, in the LPR, the surface
56
surface potential shows the same trend of decay, indicating potential decreases along with the SiO2 nanowire from top to
57
ChemNanoMat 2020, 6, 212 – 217 [Link] 213 © 2019 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
 Communication

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31 Figure 2. (a) Surface potential of SiO2 NAE as a function of oxidation time. (b) Normalized potential decay curves of SiO2 NAE with different oxidation time and
32 the potential of control samples. (c) The top view SEM image of SiO2 NAE, the corresponding (d) AFM topography image, (e) surface potential image, the
dashed line corresponds to the profile in g, (f) 3D image of surface potential image and (g) correspondi ng height profile and potential profile.
33
34
35
36
bottom. As presented in Figure 3e, the surface potential decays the charges trapped on the SiO2 nanowires are less than or
37
in the first day both in the HPR and LPR at room temperature equal to 80% while the other 20% is on the planar SiO2 layer.
38
(25 °C) and the relative humidity of 60%. In addition, the LPR is The results are confirmed by the corresponding potential profile
39
gradual converted into the HPR, which is demonstrated by the in Figure 4b. Furthermore, the potential difference between top
40
corresponding potential profile in Figure 3f. and bottom of the SiO2 nanowire is also an important feature to
41
In order to explain the different charge distribution in the determine the region in SiO2 NAE. As shown in Figure 4c, the
42
HPR and the LPR, we proposed that the charges stored in two potential difference between top and bottom of the SiO2
43
regions can be divided into two parts as depicted in Figure S3, nanowire in the HPR (UH = 0.8 V) and the LPR (UL = 2.2 V) are
44
respectively. The one part is the charges trapped on the SiO2 consistent with the simulated potential difference only when
45
nanowires while the other is the charges trapped on the planar the fraction of charge trapped on SiO2 nanowires are 90% and
46
SiO2 layer. In the HPR, the fraction of charges trapped on the 80%. However, when the fraction of charges trapped on SiO2
47
SiO2 nanowires are different from the ones in the LPR. nanoarray is 80%, the simulated potential is still higher than the
48
Furthermore, in order to quantitatively determine the fraction real potential in the LPR. As shown in Figure 4d, only when the
49
of charges trapped on the SiO2 nanowires in the two regions, total amount of the charges in the LPR is 0.4 time of that in the
50
the potential distribution of the SiO2 NAE (oxidized for 1.5 h) HPR, the simulated potential nearly matches with the real
51
was simulated by finite element method (COMSOL, detail potential distribution in the LPR. Based on these results of
52
parameter was presented in Table S1). As depicted in Figure 4a, SKPM measurement and COMSOL simulation, the charge
53
when the fraction of charges trapped on the SiO2 nanowires are distribution in the SiO2 NAE can be revealed. The charges are
54
equal or greater than 90% (the other 10% is on the planar SiO2 trapped in two regions within the SiO2 NAE after corona
55
layer), the potential distribution is similar with the HPR. Never- charging, namely the HPR and the LPR. The total amount of the
56
theless, when the potential distribution is similar with the LPR, charges stored in the LPR is 0.4 time of that in the HPR. In the
57
ChemNanoMat 2020, 6, 212 – 217 [Link] 214 © 2019 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
 Communication

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22 Figure 3. Potential distribution of SiO2 NAE in the section orientation. (a) Schematic illustration of cross section potential test based on SKPM mode. (b) The
23 cross-sectional SEM image of SiO2 NAE. (c) The cross-sectional AFM topography image of SiO2 NAE. (d) The corresponding potential image in initial stage and
(e) After a day decaying (at 25 °C and 60% RH), division of surface image are marked as white line and denoted as A, B, C, D and E in d and A1, B1, C1, D1 and E1
24 in e, and the red lined corresponds to the profile in f. (f) Corresponding potential profile.
25
26
27
HPR, 90% of the charges trapped on the SiO2 nanowires while calculation method of H2O diffusion barrier used in this work is
28
the other 10% trapped on the planar SiO2 layer. And in the LPR, based on a previous limited diffusion model.[32] In our model,
29
more than 20% of the charges retain on the planar SiO2 layer. the maximum H2O diffusion barrier in the interface of Si/SiO2
30
Based on the charge distribution in the SiO2 NAE and the can be expressed as
31
surface potential decay curves, the mechanism of charge decay
32
can be proposed as follows. Firstly, the charges trapped on the Em ¼ Ed þ DEðzÞ (1)
33
planar part of the SiO2 NAE decay rapidly in the first few days.
34 DEðzÞ ¼ 0:89 þ ð2:0 0:91ÞlðzÞ
Subsequently, the rate of charge decay becomes slow because (2)
35
almost all the charges restore on the SiO2 nanowires. Finally, a
36 zn 1 z n 1 2n lnðzÞ
stable charge distribution maintains on the SiO2 nanowires. So, lðzÞ ¼ (3)
37 ðzn þ 1Þ2 ðz n þ 1Þ2 ðzn þ 1Þðz n þ 1Þ
the charges trapped on a single SiO2 nanowire play a key role in
38
the charge stability of the SiO2 NAE. According to the previous
39
literatures, the charge decay process of the electrets is where Em, Ed, DEðzÞ, z, n and lðzÞ is the maximum H2O diffusion
40
susceptible to the external environments, such as temperature, barrier in the SiO2/Si interface, the H2O diffusion barrier in the
41
humidity, pressure and radiation, etc..[27] So, the percolation bulk SiO2, the maximum increment of H2O diffusion barrier in
42
model of H2O within the electrets is applied to explain the the SiO2/Si interface induced by compressive stress, the ratio of
43
charge decay of SiO2 electrets during the isothermal process.[28] outer radius and inner radius, n = 1/(1-v), and v is the Poisson’s
44
This model supposes that the external charges shunt the ratio of SiO2 (0.17),[29] the normalized strain energy. The 0.89 eV
45
surface and neutralize the trapped charge when the percolation is derived form the difference between the H2O diffusion barrier
46
density of H2O reaches a critical point. And it has also been (1.8 eV) in the higher density SiO2 (2.4 g/cm3)[30] and the H2O
47
reported that oxidized Si nanowires show a large H2O diffusion diffusion barrier (0.91 eV) in the bulk SiO2.[33] The H2O diffusion
48
barrier in the high density SiO2 layer near the SiO2/Si interface barrier in the maximal density SiO2 (2.7 g/cm3) can be set as
49
and the high compressive stress near the SiO2/Si interface 2.0 eV.[34]
50
increased the H2O diffusion barrier.[29–30] As the trapped charges Based on this model, the normalized strain energy and the
51
locate near the interface of SiO2/Si interface,[31] we believe that maximum H2O diffusion barrier in the interface of Si/SiO2 are
52
the large H2O diffusion barrier in the interface of Si/SiO2 can illustrated in Figure 4e and 4f. The maximum H2O diffusion
53
prevent the concentration of H2O from reaching the critical barrier in the interface of Si/SiO2 increases as the oxidation time
54
percolation density at a fast rate. To validate this mechanism, extends. Meanwhile, the maximum H2O diffusion barrier in the
55
the H2O diffusion barrier in the SiO2/Si interface for the SiO2 interface of Si/SiO2 is larger than that in the planar SiO2 (Eplanar),
56
NAE under different oxidation time was calculated. The which is derived from equation (1), (2) and (3) when the z is set
57
ChemNanoMat 2020, 6, 212 – 217 [Link] 215 © 2019 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
 Communication

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
Figure 4. (a) Series of image of COMSOL simulation of surface potential distribution of SiO2 NAE as function of fraction of charge in SiO2 nanowire structure,
35 the dashed line arrow corresponds to the potential profile in b. (b) Corresponding potential profile as indicated in a. (c) potential difference between top and
36 bottom of the SiO2 nanowire as a function of fraction of charge in SiO2 nanoarray. (d) Simulated potential profile (80% charge trapped in the SiO2 nanoarray
37 and total amount of charge is 0.4 time of that in high potential section) and real potential profile in low potential section. (e) Oxidation thickness and
normalized strain energy as a function of oxidation time. (f) Maximum H2O diffusion barrier in the interface of SiO2/Si as a function of oxidation time.
38
39
40
41
as 1.[35] However, the charge stability of the SiO2 NAE oxidized Conclusion
for 2 h is inferior to that of the SiO2 NAE oxidized for 1.5 h as
42
discussed in Figure 2e. As we have already known that the rate In summary, we report a long-term charge stable SiO2 NAE
43
of surface potential decay is highly accelerated with high without additional water repellent treatment prepared by dry
44
internal electric field due to the large charge drift rate.[21] oxidizing Si nanoarray. The optimum charge stability SiO2 NAE
45
Although the SiO2 NAE oxidized for 1.5 h and 2 h almost have is obtained when the oxidation time is 1.5 h. The potential
46
the same thickness (as shown in the Figure 4c), the surface decay of the SiO2 NAE (oxidized for 1.5 h) is less than 46%
47
potential of the SiO2 NAE (oxidized for 2 h) is higher than the during 60 days with an original potential of 120 V. The
48
one oxidized for 1.5 h. Therefore, because of larger charge drift excellent charge stability is ascribed to its unique charge
49
rate in the stronger internal electric field, the SiO2 NAE (oxidized distribution and the large H2O diffusion barrier in the interface
50
for 2 h) has the poor charge stability. It has to point out that of Si/SiO2. Two kinds of charge distribution exist in the SiO2
51
the variation of charge trap defects is not discussed in this NAE. The one is the 90% of charge trapped in the SiO2
52
work. Notwithstanding its limitation, this article does prove the nanoarray (denoted as chargeIdistribution) corresponding to
53
excellent charge stability of the SiO2 NAE and gives an insight the H-potential distribution. The other is the 80% of charge
54
view of charge distribution in nanoscale for the SiO2 NAE, which trapped in the SiO2 nanoarray corresponding to the L-potential
55
will greatly benefit the design and development of high- distribution (denoted as II charge distribution). Additionally,
56
performance MEMS devices based on the electrets. total amount of charge in LPR is 0.4 time of that the HPR. In the
57
ChemNanoMat 2020, 6, 212 – 217 [Link] 216 © 2019 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
 Communication
initial charge decay process, the II charge distribution can be Keywords: electret · atomic force microscopy · scanning Kelvin
1
converted to the I charge distribution, and then the residual probe microscopy · charge distribution · charge stability
2
charge trapped in the SiO2 nanoarray. Finally, the large H2O
3
diffusion barrier in the interface of Si/SiO2 effectively hinders
4 [1] Y. Suzuki, Electr. Electron. Eng. 2011, 6, 101–111.
the charge decay. Therefore, this SiO2 NAE can be applied to
5 [2] B. Wang, J. Zhong, Q. Zhong, N. Wu, X. Cheng, W. Li, K. Liu, L. Huang, B.
fabricate new type and high-performance electret-based devi- Hu, J. Zhou, Adv. Electron. Mater. 2016, 2, 1500408.
6
ces and MEMS sensors for its long-term charge stability and [3] K. Hagiwara, M. Goto, Y. Iguchi, T. Tajima, Y. Yasuno, H. Kodama, K.
7 Kidokoro, Y. Suzuki, IEEE Trans. Dielectr. Electr. Insul. 2012, 19, 1291–
unique structure.
8 1298.
9 [4] D. Hohm, R. Gerhard-Multhaupt, J. Acoust. Soc. Am. 1984, 75, 1297–
1298.
10 [5] H. Amjadi, G. M. Sessler, In Conference on Electrical Insulation & Dielectric
Supporting Information Phenomena 1998, pp. 132–135.
11
12 [6] S. A. H. Sabzevari, M. Moavenian, Ultrasonics. 2017, 73, 114–124.
Supporting Information is available from the Wiley Online [7] U. Mescheder, P. Urbanovic, B. Muller, S. Baborie, In Proc. PowerMEMS
13 2008, Sendai, Japan, Nov 2008, pp. 501–504.
Library or from the author.
14 [8] P. Günther, Z. Xia, J. Appl. Phys. 1993, 74, 7269–7274.
15 [9] F. Wang, O. Hansen, IEEE Electron Device Lett. 2013, 34, 1047–1049.
[10] M. Suzuki, T. Wada, T. Takahashi, T. Nishida, Y. Yoshikawa, S. Aoyagi, In
16 2014 IEEE 27th Int. Conf. Micro Electro Mech. Syst, California, America,
Experiment Section Jan 2014, pp. 417–420.
17
18 Synthesis of Si nanoarray and SiO2 nanoarray: The Si nanoarray were [11] Y. C. Chen, K. Y. Song, K. Morimoto, Y. Suzuki, In 2015 IEEE 28th Int.
synthesized via a Ag-assisted chemical etching method.[22] The p Conf. Micro Electro Mech. Syst, Estoril, Portugal, Jan 2015, pp. 1086–
19 1089.
type (100) Si wafer (R < 0.05 Ω cm) was used in the experiment.
20 [12] Z. Y. Zhou, N. Tian, J. T. Li, I. Broadwell, S. G. Sun, Chem. Soc. Rev. 2011,
First, the cleaning of Si wafer with acetone, ethanol, deionized 40, 4167–4185.
21 water and HF was conducted. Then, the (20 nm) Ag films were [13] S. V. Kuchibhatla, A. Karakoti, D. Bera, S. Seal, Prog. Mater. Sci. 2007, 52,
22 deposited via an e-beam evaporator (Denton, Explore 14) with the 699–913.
23 rate of 0.8 nm/s. In the next, the Si wafer covered with Ag were [14] E. Roduner, Chem. Soc. Rev. 2006, 35, 583–592.
24 immersed into a mixed solution of HF, H2O2 and H2O (volume [15] H. P. Wang, D. Periyanagounder, A. C. Li, J. H. He, IEEE Access. 2018, 7,
respectively is 8.3, 0.5 and 40 mL) in a Teflon bottle and reacted for 19395–19400.
25 [16] J. R. D. Retamal, C. H. Ho, K. T. Tsai, J. J. Ke, J. H. He, IEEE Trans. Electron
a certain time at room temperature. Finally, the etched sample
26 Devices 2019, 66, 938–943.
were rinsed in deionized water and dried. Subsequently, the as [17] Z. Zhang, Y. Ning, X. Fang, J. Mater. Chem. C. 2019, 7, 223–229.
27 prepared Si nanoarray was oxidized in a controlled atmosphere [18] S. Das, M. J. Hossain, S. F. Leung, A. Lenox, Y. Jung, K. Davis, J. H. He, T.
28 furnace at 1000 °C with a volume flow rate of oxygen 100 sccm for Roy, Nano Energy. 2019, 58, 47–56.
29 different oxidation time to obtain the SiO2 nanoarray. [19] Y. Ning, Z. Zhang, F. Teng, X. Fang, Small. 2018, 14, 1703754.
30 [20] A. Mellinger, F. C. González, R. Gerhard-Multhaupt, IEEE Trans. Dielectr.
SKPM testing of SiO2 NAE: The surface potential distribution was Electr. Insul. 2004, 11, 218–226.
31 measured by a MFP-3D atomic force microscope in the SKPM mode [21] D. Rychkov, A. Kuznetsov, A. Rychkov, IEEE Trans. Dielectr. Electr. Insul.
32 at room temperature with a relative humidity of 60%. A conductive 2011, 18, 8–14.
33 Si based probe (coating with Pt, ARROW-EFM-10 from Nanoworld) [22] C. Chartier, S. Bastide, C. Lévy-Clément, Electrochim. Acta 2008, 53,
was used in the tap mode (amplitude 520 mV) and SKPM mode 5509–5516.
34 [23] X. Zou, J. Zhang, Sci. China Technol. Sci. 2011, 54, 2123–2129.
(above the surface 300 nm) to obtain the topography and potential
35 [24] B. Du, Z. Yang, Z. Li, J. Li, IEEE Trans. Dielectr. Electr. Insul. 2017, 24,
data of SiO2 NAE. For the potential test along the direction of 1340–1348.
36 orientation of nanoarray, the charged sample firstly were cut by [25] V. Leonov, C. Van Hoof, M. Goedbloed, R. Van Schaijk, IEEE Trans.
37 diamond saw, and then the vertical sample was fixed and placed in Dielectr. Electr. Insul. 2012, 19, 1253–1260.
38 test substrate of AFM. [26] S. Sahli, A. Bellel, Z. Ziari, A. Kahlouche, Y. Segui, J. Electrostat. 2003, 57,
39 169-181.
Characterization: The SEM image was taken by a scanning electron [27] V. N. Kestelman, L. S. Pinchuk, V. A. Goldade, Electrets in Engineering:
40 microscopy (SEM, SU8020). Transmission electron microscopy (TEM, Fundamentals and Applications, Springer-Verlag New York Inc, 2013, pp.
41 TECNAI F20) and EDX elemental mapping (TECNAI F20) were used 24–27.
42 to determine the morphology and element distribution. The surface [28] A. Viraneva, T. Yovcheva, E. Gencheva, G. Mekishev, J. Phys. Conf. Ser.
potential of the samples was measured by an electrostatic 2010, 253, 012069.
43 [29] H. Ohta, T. Watanabe, I. Ohdomari, Jpn. J. Appl. Phys. 2007, 46, 3277.
voltmeter (ISOPROBE, 279).
44 [30] T. Bakos, S. N. Rashkeev, S. T. Pantelides, Phys. Rev. Lett. 2002, 88,
45 055508.
[31] W. Olthuis, P. Bergveld, IEEE Trans. Dielectr. Electr. Insul. 1992, 27, 691–
46 697.
Acknowledgement
47 [32] H. Cui, C. Wang, G. Yang, Nano Lett. 2008, 8, 2731–2737.
48 [33] H. Wakabayashi, M. Tomozawa, J. Am. Ceram. Soc. 1989, 72, 1850–1855.
Fei Liang and Hua Yang Li contribute equally to this work. This [34] P. Koziatek, J. Barrat, D. Rodney, J. Non. Cryst. Solids. 2015, 414, 7–15.
49 [35] H. Liu, D. Biegelsen, F. Ponce, N. Johnson, R. Pease, Appl. Phys. Lett.
research was supported by the National Key R & D Project from
50 1994, 64, 1383–1385.
Ministry of Science and Technology, China (Grant Nos.
51
2016YFA0202701 and 2016YFA0202703), National Science
52
Foundation of China (Grant No. 51572030), Natural Science
53 Manuscript received: October 24, 2019
Foundation of Beijing Municipality (Grant No. 2162047), and
54 Accepted manuscript online: November 11, 2019
China Thousand Talents Program. Version of record online: January 14, 2020
55
56
57
ChemNanoMat 2020, 6, 212 – 217 [Link] 217 © 2019 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim