Electric-Field Enhanced Ni Induced Lateral Crystallization Rate

Subscriber access provided by UNIV OF WISCONSIN OSHKOSH

Article

Electric-field enhanced Ni induced lateral crystallization rate saturation in heavily phosphorus-doped amorphous silicon Ji-Su Anh, Deok-kee Kim, and Seoung-Ki Joo Cryst. Growth Des., Just Accepted Manuscript • DOI: 10.1021/acs.cgd.5b01158 • Publication Date (Web): 08 Dec 2015 Downloaded from http://pubs.acs.org on December 9, 2015

Just Accepted “Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted online prior to technical editing, formatting for publication and author proofing. The American Chemical Society provides “Just Accepted” as a free service to the research community to expedite the dissemination of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts appear in full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been fully peer reviewed, but should not be considered the official version of record. They are accessible to all readers and citable by the Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered to authors. Therefore, the “Just Accepted” Web site may not include all articles that will be published in the journal. After a manuscript is technically edited and formatted, it will be removed from the “Just Accepted” Web site and published as an ASAP article. Note that technical editing may introduce minor changes to the manuscript text and/or graphics which could affect content, and all legal disclaimers and ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors or consequences arising from the use of information contained in these “Just Accepted” manuscripts.

Crystal Growth & Design is published by the American Chemical Society. 1155 Sixteenth Street N.W., Washington, DC 20036 Published by American Chemical Society. Copyright © American Chemical Society. However, no copyright claim is made to original U.S. Government works, or works produced by employees of any Commonwealth realm Crown government in the course of their duties.

Page 1 of 29

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 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Crystal Growth & Design

Electric-field enhanced Ni induced lateral crystallization rate saturation in heavily phosphorus-doped amorphous silicon

Ji-Su Ahn1, Deok-kee Kim1,*, and Seung-Ki Joo2

1

Nano Device Laboratory, Sejong University, 98 Gunja-dong, Gwangjin-gu, Seoul 143747, Korea

2

School of Materials Science and Engineering, Seoul National University, San 56-1, Shinrim-Dong, Kwanak-gu, Seoul 151-742, Korea *[email protected]

1

ACS Paragon Plus Environment


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 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 2 of 29

Abstract

Electric-field enhanced Ni induced lateral crystallization rate saturation in ndoped amorphous silicon thin films was investigated as a function of PH3 doping time. Crystallization rate decreased as the doping time increased and saturated at 2-3 µm/hr after 3 min doping time, irrespective of an externally applied ɛ-field, which was attributed to a much higher internal ɛ-field by the space charge at the interface due to heavy n-type doping. Not only crystallization rate but also microstructural changes were dependent on the doping time and the direction of the ɛ-field. The microstructure of the n-doped samples with ɛ-field showed two different types of bi-directional needle network structures with different secondary growth directions, which implies that the microstructures have different preferred orientations with different crystallization mechanisms involved. A mechanism describing the crystallization rate saturation was proposed employing the field dependent charged vacancy migration and internal interfacial field.

Key words: Polycrystalline silicon, Microstructure, MILC, Electric Field, Doping time

2


Page 3 of 29

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 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


1. Introduction Polycrystalline silicon thin films have widely been applied in thin film transistors, solar cells, high-resolution liquid crystal displays, light-emitting diodes, and sensors [1-3].

Polycrystalline Si thin films are frequently fabricated by crystallizing

amorphous Si (a-Si) thin films. A variety of methods for lowering the crystallization temperature of a-Si thin films have been developed. Excimer laser annealing [4] is one of the promising ways to achieve large grain size poly Si thin films at low temperatures, but has the drawbacks of high costs and nonuniform grain size. Solid phase crystallization (SPC) [5-8] requires a high-temperature process and many substrates, including most forms of glass, where all of the processing steps need to be limited to temperatures below 550 °C, cannot withstand the thermal processing. Rapid thermal annealing (RTA) uses infrared radiation as a heating source that does not heat the glass substrate since glass is transparent to the infrared radiation. RTA has the advantage of the high heating speed (up to 60 °C/s) that reduces the crystallization time. However, the obtained grain size is in the range of a few micrometers. Metal-induced crystallization (MIC) [9-12] has been investigated as an

3



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 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 4 of 29

alternative crystallization process for thin-film device fabrication. MIC uses metals like aluminum, nickel, gold, silver, palladium, etc. to lower the crystallization temperature compared to SPC. Among the various metals used for the study of the MIC [13-16], Ni has been the preferred choice with its low residual metal contaminants in the poly-Si region [17]. In order to further reduce the leakage current caused by the metal agents, metal-induced lateral crystallization (MILC), where the crystal growth occurs in the lateral direction from a metal electrode, has been introduced. The effect of electric field on the growth and microstructure in doped amorphous silicon thin films during Ni induced lateral crystallization in view of doping time was reported previously [18]. However, it was not investigated in detail how different level of dopant concentration in Si affects the rate of MILC process and consequently the crystal quality of Si films resulting from MILC. Ni induced lateral crystallization rate saturation in n-doped a-Si thin films as a function of n-type doping time has not been reported as far as we know. In this work, we report the MILC growth rate saturation of a-Si thin films under ɛ-field due to heavy phosphorus doping in a-Si thin films. The MILC growth rate and microstructural changes were dependent on the doping time and the direction of the ɛfield. The doping time dependent MILC growth rate saturation of phosphorus-doped a-

4


Page 5 of 29

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 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


Si thin films was explained employing the previously proposed field dependent charged vacancy migration [18] and internal field induced by heavy doping.

2. Experimental section

50-nm-thick a-Si thin films were prepared by the same low-pressure chemical vapor deposition using disilane (Si2H6) as the Si source gas, which was described in our previous report [18]. After the a-Si deposition, the samples were doped with PH3 (ntype) for varying times (0-5 min) by ion mass doping (IMD). By varying the doping time, the concentration of n-type dopants, which can be inferred from the current measurements in Fig. 1, was varied. Subsequently, 200-nm-thick Pt electrode and 5-nm-thick Ni island patterns were formed by the same sputter-deposition and lift-off process which was described in our previous report [18] and as shown schematically in the inset of Fig. 1(b). The areas of the sample and Ni islands were 1 cm X 2 cm and 200 µm X 200 µm, respectively. The samples were annealed at 550 °C for 5 hrs in a vacuum tube furnace to induce crystallization of a-Si while an ɛ-field was applied. Ɛ-field with a magnitude of 100 V/cm was uniformly applied using a Keithley 237 power source between the anode and

5



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 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 6 of 29

the cathode, and the current measurements were done while the sample was being annealed in the vacuum tube furnace using a Keithley 2000 multi-meter. The microstructure of the samples was analyzed by optical microscopy (OM) and field-emission scanning electron microscopy (FE–SEM). The samples were etched using Secco-solution before the FE-SEM observation to reveal the a-Si and c-Si interface and the grain boundaries in c-Si.

3. Results and discussion

Fig. 1(a) shows the current through the samples doped with PH3, for varying times (0-5 min), during the thermal annealing at 550 °C for 5 hrs while an ɛ-field of 100 V/cm was applied. In the sample doped with PH3 for 0 to 2 mins, the current did not increase much during the 5 hr anneal at 550 °C, while in the sample doped with PH3 for 3 to 5 mins, the current increased significantly during the anneal. An exemplary current measurement of intrinsic a-Si including the ramp-up and ramp-down steps during annealing at 550 °C for 2 hrs is shown in Fig. S1. The current before the ramp-up was extremely low. With the increased temperature during the ramp-up, the current in intrinsic a-Si increased significantly by thermal activation of carriers. When the sample

6


Page 7 of 29

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 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


is cooled down to room temperature after annealing, the current sharply decreased without the thermal activation. Fig. 1(b) shows the initial and final current values as the IMD doping time increased from 0 to 5 min, which was extracted from the current measurements in Fig. 1(a). The initial and final current increased as the doping time increased, where the rate of increase was higher after 3 min doping. The increase in the initial current is attributed to the higher doping as the doping time increased, while that in the final current is attributed to the MIC and MILC growth, and the dopant activation in the crystallized silicon by the increased doping. The significant increase of current during 550 °C annealing in the case of 4 or 5 min doping time is believed to be due to the increased activation with sufficient phosphorous dopants in a-Si films, where dopants are only partially activated in a-Si compared to c-Si [19]. The calculated resistivity data corresponding to the initial and final current measurements in Fig. 1(b) are shown in Fig. S2. Fig. 2 shows OM images of MIC, MILC, and a-Si region of the sample doped with PH3 for (a) 0, (b, c) 1, and (d) 5 mins after annealing at 550 °C for 5 hrs, respectively. Figs. 2(a), 2(c), and 2(d) correspond to the samples with the applied ɛ-field of 100 V/cm, while Fig. 2(b) corresponds to the sample without the ɛ-field. OM images

7



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 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 8 of 29

show that the region where the Ni thin film had been selectively deposited was crystallized by MIC, and the region surrounding the deposited nickel thin film was laterally crystallized by MILC from several to several tens of micron-meters [18]. OM images in Fig. 2 show that interface between MILC and a-Si region looked flat for the undoped sample, while the interface looked rugged for doped samples, irrespective of ɛfield. The rugged interface for n-type a-Si seems to be due to the fact that there are many competing driving forces for MILC growth for n-type a-Si such as the attractive force between the charged Ni- ion and the activated dopants, the interaction between the negatively charged Ni- ions and the applied ɛ-field, and the internal interfacial ɛ-field, which will be explained in detail in the following. Depending on the doping and the applied ɛ-field, the resulting microstructure and MILC growth rate of a-Si thin films varied. OM images in Figs. 2(c) and 2(d), corresponding to 1 and 5 min doping time, showed that the microstructure were similar, although the MILC growth rate was suppressed for 5 min doping time. Fig. 3(a) shows FE-SEM images of MILC and a-Si region of the sample doped with PH3 for 1 min after annealing at 550 °C for 5 hrs without ɛ-field, while Figs. 3(b) and 3(c) show those with ɛ-field in (b) the cathode and (c) the anode directions. The morphology and microstructure of MILC and a-Si region of the sample were

8


Page 9 of 29

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 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


investigated using the FE-SEM images in Figs. 3(a)-3(c). The locations where the FESEM images in Figs. 3(a)-3(c) were taken are marked in the OM images in Figs. 2(b) and 2(c). The insets in Figs. 3(a)-3(c) show the corresponding magnified FE-SEM images of MIC, MILC, and a-Si region of the sample. Table I summarizes the expected preferred orientation and characteristic microstructure of the MILC growth region for the samples doped with PH3 for 1 min with and without the electric field in both the anode and the cathode directions, respectively, in Figs. 3(a)-3(c). The maximum length of the crystallites in Table I was measured for an area of 10 µm X 10 µm. The microstructure of the MILC region of the sample doped with PH3 without the ɛ-field showed a multi-directional needle network structure of branched crystallites whose primary and secondary growth directions have the regular angles of 70 or 110 degrees which is expected to be related with a preferred growth orientation of direction in (110) plane [9]. The maximum length of crystallite was over 5.6 µm, while the width was 90 - 100 nm. The microstructure of the sample doped with PH3 with the ɛ-field facing the anode had a bi-directional needle network structure of branched crystallites whose primary and secondary growth directions had the regular angles of 85 or 95 degrees, which is expected to be related with a preferred growth orientation of direction in (100) plane. Slight deviation from 90 degree for normal direction in

9



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 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 10 of 29

(100) plane seems to be due to the presence of the electric field. This showed different angles from those (70 or 110 degrees) in the sample without the ɛ-field. Further detailed study of the microstructure with the MILC growth mechanism depending on the electric field is beyond the scope of the present study. The maximum length of crystallite was 1.6 µm, while the width was 30 - 40 nm. The microstructure of the MILC region of the sample doped with PH3 with the ɛ-field facing the cathode had a bi-directional needle network structure with enhanced directionality whose primary and secondary growth directions had the regular angles of 70 or 110 degrees, which is expected to be related with a preferred growth orientation of direction in (110) plane [9]. The maximum length of crystallite was 4.5 µm with the enhanced directionality, while the width was 50 - 60 nm, which is thinner than that (90 - 100 nm) of MILC without the ɛ-field. Fig. 3(d) shows the MILC growth rate of PH3-doped a-Si thin films as a function of doping time (0-5 mins) in the cathode (-, red) and the anode (+, blue) directions after 5 hr annealing at 550 °C with an ɛ-field of 100 V/cm, together with the MILC growth rate without ɛ-field (black). Without the ɛ-field, MILC growth rate decreased almost linearly from 7.38 to 1.86 µm/hr as the doping time increased from 0 (undoped) to 3 min. After 3 mins, the MILC growth rate saturated around 2 µm/hr. With the ɛ-field in the anode direction, MILC growth rate was 14.54 µm/hr for the undoped sample and

10


Page 11 of 29

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 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


decreased to 2.92 µm/hr after phosphorus doping for 3 mins. For higher than 3 min doping, the MILC growth rate was saturated around 3 µm/hr. On the other hand, with the ɛ-field in the cathode direction, MILC growth rate was 5.78 µm/hr for the undoped sample and decreased to 2 µm/hr after phosphorus doping for 3 mins. After 3 min doping, the MILC growth rate was saturated around 2 µm/hr and seemed to saturate for both the samples with and without an applied ɛ-field. The decrease of MILC growth rate for phosphorus doped a-Si thin films without ɛ-field can be explained using the previously proposed 3 step Ni ion and Ni vacancy hopping mechanism [18]. Detailed atomic arrangements of Ni and Si atoms during MILC were also proposed in our previous work [20, 21]. Ni atoms in NiSi2 phase are negatively charged because of the negative Mulliken charge [22]. Because of the attractive force between the charged Ni- ion (or Ni+v vacancy) and the activated dopants (P+), the velocity of the c-Si/NiSi2 interface would be slowed down so that MILC growth rate would be decelerated by the increased n-type doping, which matched with the results in Fig. 3(d). The rate and the percentage of activation of P+ dopants in crystallized region are believed to be much higher than those in a-Si region.

During

MILC, vigorous atomic movements occur in the crystallized region, which should facilitate the activation of P+ dopants, while the dopants are only partially activated in

11



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 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 12 of 29

the a-Si region [19]. The interaction between the negatively charged Ni- ions and the applied ɛ-field can be used to explain partly why phosphorus-doped a-Si showed higher and lower MILC growth rate in the anode and cathode directions, compared to that without ɛ-field, respectively. The applied ɛ-field drifts the charged Ni- ions toward the leading edge in the anode direction, which promotes the MILC growth rate in the anode direction. Likewise, the drift in the opposite direction retards the MILC growth in the cathode direction with the same mechanism which we proposed for ɛ-field enhanced MILC without doping [18]. In addition to the charged Ni- ions, in heavily doped silicon, the concentration of the charged vacancies is known to increase, which can affect the MILC growth rate in the anode and cathode directions. Concentration changes of charged vacancies (V2-, V-, V0, V+) in crystalline silicon [18] can also be used to explain why phosphorus-doped a-Si showed higher and lower MILC growth rate in the anode and cathode directions, respectively, which is shown in schematic diagrams of Ni ion and Ni vacancy hopping model (Fig. 4). The overall concentration of charged vacancies in the doped silicon is known to increase or decrease significantly (from tens to tens of thousands times) compared to that of intrinsic silicon depending on the temperature (T) and the Fermi energy level (EF) [23].

12


Page 13 of 29

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 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


For the n-doped sample, V- and V2- vacancies are dominant so that the vacancy concentration increases and decreases in the anode and cathode directions, respectively, by the migration of the negatively charged vacancies due to the ɛ-field.

This makes

the growth rate accelerated and decelerated in the anode and cathode directions, respectively, by promoting and demoting the Si atom adsorption and the atomic rearrangement of Si [18]. Fig. 4 shows schematic diagrams of Ni ion and Ni vacancy hopping model explaining the MILC growth rate of the sample doped with PH3 in (a) the anode and (b) the cathode directions by employing the charged vacancy migration model and internal interfacial electric field. MILC rate saturation after 3 min doping in Fig. 3(d) can be explained using internal interfacial ɛ-field. For the samples doped with PH3 for more than 3 mins, the crystallized region of the Si thin films are expected to be heavily ndoped. The non-crystallized region (a-Si) has a high density of states in the mobility gap, some of which are associated with topological defects such as dangling bonds and multi vacancies.

These defect states can take an extra electron and thus can be thought as

acceptors.

In a simplified approach, the amorphous non-crystallized region can be

assumed as a material with a density nx of acceptor states, and the heterojunction at the interface can be treated as a p-n junction [24]. The displacement of the Fermi level due

13



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 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 14 of 29

to doping in the crystallized region leads to a charge transfer at the junction. In the case of n-type doping, some of the defects in the amorphous region can accept an electron, and their migration towards the interface where subsequent crystallization takes place during the MILC will become easier or harder by the ɛ-field. The width of the space charge layer and the internal ɛ-field at the interface can be calculated using the following formulae [25]:

W =

ξ =

2 ε s (1 + n x /n )V bi

q

nx

n xV bi , ε s (1 + n x / n )

2q

(4)

(5)

where n is the dopant concentration in the crystallized region, nx is the concentration of defects in a-Si which plays the part of acceptors, Vbi is the built in potential, and ɛs is the dielectric permittivity of Si. With the doping level of 1016 - 1020 /cm3 and Vbi~Eg/2, the typical values of the width of the space charge layer and the ɛ-field is found to be W ~ 38 - 3.8 nm and ξ ~ 3x104 V/cm - 3 MV/cm, respectively. The internal interfacial ɛ-field of 3x104 V/cm - 3 MV/cm is much higher than the externally applied ɛ-field of 100 V/cm, which makes the effect of the externally applied ɛ-field minimal and the MILC growth rate saturated in the heavily n-doped samples, regardless of ɛ-field. As the doping level increases, MILC rate seems to saturate since higher order interfacial

14


Page 15 of 29

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 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


internal field affects more portion of the total MILC growth region.

4. Conclusions

MILC growth rate saturation of a-Si thin films under ɛ-field due to heavy phosphorus doping was investigated.

The MILC growth rate and microstructure

varied depending on doping time and ɛ-field. The microstructure of the n-doped samples with ɛ-field showed two different types of bi-directional needle network structures with different secondary growth direction. The angle between the primary and secondary growth directions were different depending on the direction of the ɛ-field in n-doped samples, which implies that the microstructures have different preferred orientations with different MILC growth mechanisms involved. For both the samples with and without an applied ɛ-field, MILC growth rate decreased as the doping time increased and saturated at 2-3 µm/hr after 3 min. MILC growth rate saturation with and without ɛ-field was explained by employing the charged vacancy migration model and taking into account the much higher internal ɛ-field by the space charge at the interface due to the heavy n-type doping, compared to the external ɛfield. For the heavily n-doped samples, the increased concentration of charged defects

15



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 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 16 of 29

will be attracted toward the interface because of their state of charge and the interfacial ɛ-field. The charged defects’ movements will be dominated by the higher internal interfacial ɛ-field in heavily n-type doped samples, which induces the MILC growth rate saturation.

Acknowledgements

This research was supported by Nano Material Technology Development Program (2015M3A7B7045496) and Indo-Korea Joint Program of Cooperation in Science and Technology (2014K1A3A1A19067299) through the National Research Foundation of Korea (NRF) of Korea funded by the Ministry of Science, ICT & Future Planning.

16


Page 17 of 29

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 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


References (1) Schropp, R.E.I.; Zeman, M. Amorphous and Microcrystalline Silicon Solar Cells: Modeling, Materials and Device Technology, Kluwer Academic Publishers 1998, Boston. (2) Choi, J.H.; Cheon, J.H.; Kim, S.K.; Jang, J. Giant-grain silicon (GGS) and its application to stable thin film transistor Displays 2005, 26, 137. (3) Mahan, A.H.; Ahrenkiel, S.P.; Schropp, R.E.I.; Li, H.; Ginley, D.S. A comparison of grain nucleation and grain growth during crystallization of HWCVD and PECVD Si:H films Thin Solid Films 2008, 516, 529. (4) Morita Y.; Noguchi, T. UV Pulsed Laser Annealing of Si+ Implanted Silicon Film and Low-Temperature Super-Thin Film Transistors Jpn. J. Appl. Phys. 1989, 28, 309. (5) Iverson, R. B.; Reif, R. J. Recrystallization of Amorphized Polycrystalline Silicon Films on SiO2: Temperature Dependence of the Crystallization Parameters Appl. Phys. 1987, 62, 1675. (6) Hatalis M. K.; Greve, D. W. Large Grain Polycrystalline Silicon by Low‐Temperature Annealing of Low‐Pressure Chemical Vapor Deposited Amorphous Silicon Films J. Appl. Phys. 1988, 63, 2260. (7) Nakazawa, K. Recrystallization of Amorphous Silicon Films Deposited by

17



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 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 18 of 29

Low‐Pressure Chemical Vapor Deposition from Si2H6 Gas J. Appl. Phys. 1991, 69, 1703. (8) Aoyama, T.; Kawachi, G.; Konishi, N.; Suzuki, T.; Okajima, Y.; Miyata, K. Crystallization of LPCVD Silicon Films by Low Temperature Annealing J. Electrochem. Soc. 1989, 136, 1169. (9) Lee S. W.; Joo, S. K. Low Temperature Poly-Si Thin-Film Transistor Fabrication by Metal-Induced Lateral Crystallization IEEE Elect. Dev. Lett. 1996, 17, 160. (10) Miyasaka M.; Stoemenos, J. Excimer Laser Annealing of Amorphous and SolidPhase-Crystallized Silicon Films J. Appl. Phys. 1999, 86, 5556. (11) Kang, I.; Han, S.; Joo, S. Novel Offset-Gated Bottom Gate Poly-Si TFTs With a Combination Structure of Ultrathin Channel and Raised Source/Drain IEEE Elect. Dev. Lett. 2008, 29, 232. (12) Lee, K.-Y. Super Grain Silicon Technology for AMOLED TV Application SID Symp. Dig. Tech. Papers, vol. 2011, 42, 175. (13) Jin, Z.; Bhat, G. A.; Yeung, M.; Kwok, H. S.; Wong, M. Nickel Induced Crystallization of Amorphous Silicon Thin Films J. Appl. Phys. 1998, 84, 194. (14) Kumar, K.U.M.; Brahma, R.; Krishna, M.G.; Bhatnagar, A.K.; Dalba, G. An optical study of Ni induced crystallization of a-Si thin films J. Phys. Condens. Matter 2007, 19, 496208.

18


Page 19 of 29

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 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


(15) Kumar, K.U.M.; Krishna, M.G. Chromium-induced nanocrystallization of a-Si thin films into the wurtzite structure J. Nanomater. 2008, 736534-1. (16) Mohiddon, M.A.; Krishna, M.G. Nanocrystalline wurtzite Si-nickel silicide composite thin films with large band gap and high resistivity J Mater. Sci. 2011, 46, 2672. (17) Wang, M.; Wong, M. Characterization of an individual grain boundary in metal induced laterally crystallized polycrystalline silicon thin-film device IEEE Trans. Electron Devices 2001, 48, 1655. (18) Ahn, J.-S.; Kim, D. K.; Seung-Ki Joo. Effect of electric field on doped amorphous silicon thin films during Ni induced lateral crystallization Materials Letters 2013, 93, 233. (19) Alpuim, P.; Chu V.; Conde J. P. Doping of amorphous and microcrystalline silicon films deposited at low substrate temperatures by hot-wire chemical vapor deposition J. Vac. Sci. Technol. A 2001, 19, 2328. (20) Ahn, J.-S.; Kim, D.; Joo, S.-K. Effect of Electric Field on Amorphous Silicon Thin Films during Ni Induced Lateral Crystallization J. Appl. Phys. 2013, 113, 223509. (21) Ahn, J.-S.; Yoon, Y.-G.; Kim, D.; Joo, S.-K. Effect of MSi2/Si(111) (M=Co, Ni) Interface Structure on Metal Induced Lateral Crystallization Thin Solid Films 2013, 542,

19



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 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 20 of 29

426. (22) Cherns, D.; Anstis G. R.;Hutchison, J. L. Atomic Structure of the NiSi2/(111)Si Interface Phil. Mag. A 1982, 46, 849. (23) Van Vechten J.A.;Thurmond, C.D. Comparison of Theory with Quenching Experiments for the Entropy and Enthalpy of Vacancy Formation in Si and Ge Phys. Rev. B 1976, 14, 3551. (24) Licoppe C.; Nissim, Y. I. Impurity-Induced Enhancement of the Growth Rate of Amorphized Silicon during Solid-Phase Expitaxy: A Free-Carrier Effect J. Appl. Phys. 1986, 59, 432. (25) Sze, S.M. Physics of Semiconductor Devices (Wiley, New York, 1981).

20


Page 21 of 29

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 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


Figures Figure 1 (a) Current through the samples doped with PH3, for varying times (0-5 min), during the thermal annealing at 550 °C for 5 hrs while an ɛ-field of 100 V/cm was applied. (b) Initial and final current values as the IMD doping time increased from 0 to 5 min. Inset shows the schematic diagram of the sample with Pt electrodes at the edges. Figure 2 OM images of MIC, MILC, and a-Si region of the sample doped with PH3 for (a) 0, (b, c) 1, and (d) 4 mins after annealing at 550 °C for 5 hrs. (a), (c) and (d) correspond to the samples with ɛ-field, while (b) to that without the ɛ-field. Figure 3 (a) FE-SEM image of MILC and a-Si region of the sample doped with PH3 for 1 min after annealing at 550 °C for 5 hrs without ɛ-field, with the corresponding magnified image in the inset. FE-SEM images of the sample in (b) the cathode and (c) the anode directions, with the corresponding magnified images in the insets. (d) MILC growth rate of PH3-doped a-Si thin films as a function of doping time (0-5 mins) in the cathode (-, red) and the anode (+, blue) directions after 5 hr annealing at 550 °C with an ɛ-field of 100 V/cm, together with MILC growth rate without ɛ-field (black). Figure 4 Schematic diagrams of Ni ion and Ni vacancy hopping model explaining the MILC growth rate of the sample doped with PH3 with an external ɛ-field in (a) the anode and (b) the cathode directions by employing the charged vacancy migration

21



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 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 22 of 29

model and internal interfacial ɛ-field.

22


Page 23 of 29

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 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


"For Table of Contents Use Only" Title: Electric-field enhanced Ni induced lateral crystallization rate saturation in heavily phosphorus-doped amorphous silicon Authors: Ji-Su Ahn, Deok-kee Kim, and Seung-Ki Joo .

Synopsis Crystallization rate decreased as the doping time increased and saturated at 2-3 µm/hr after 3 min doping time, irrespective of an externally applied ɛ-field, which was attributed to a much higher internal ɛ-field by the space charge at the interface due to heavy n-type doping. A mechanism describing the crystallization rate saturation was proposed employing the field dependent charged vacancy migration and internal interfacial field.

23



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 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 24 of 29

Supporting Information

Figure S1 An exemplary current measurement of intrinsic a-Si including the ramp-up and ramp-down steps during annealing at 550 °C for 2 hrs. Figure S2 Calculated resistivity data corresponding to the initial and final current measurements in Fig. 1(b).

24


Page 25 of 29

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 35 36 37 38 39 40 41 42


Table I. Preferred Orientation and microstructure of MILC growth region for the sample doped with PH3 for 1 min Case Without E-Field

Dominant Microstructure Multi-directional (~70 º) needle network

Expected Preferred Maximum Length of Crystallite (μm) Orientation

With E-Field, anode

bi-directional (~85 º)needle network

/(110) /(100)

With E-Field, cathode

bi-directional (~70 º) needle network (with enhanced directionality)

/(110)


>5.6 1.6 4.5

Width of Needle Crystallite (nm) 90~100 30~40 50~60


3 3

2 1

Current (mA)

Current (mA)

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 35 36 37 38 39 40 41 42

Page 26 of 29

PH3 IMD 2 min PH3 IMD 1 min

0

Intrinsic

0

1

2

3

4

5

2

Initial

1 0

Time (hr)

Final

0

1

2

3

Doping time (min)

(a)

(b) Fig. 1 ACS Paragon Plus Environment

4

5

Page 27 of 29

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 35 36 37 38 39 40 41 42


①

30 µm

MIC

v-

MIC

v+

Undoped v-: 5.8 µm/hr

(-)

v+: 14.5 µm/hr

v

(+)

PH3 IMD 1 min

MILC Without electric field

(a) MIC

v- PH3 IMD 1 min

②

v-: 4.3 µm/hr v+: 8.5 µm/hr

v

v: 5.2 µm/hr

MILC

With electric field

(-)

30 µm

(b) 30 µm

MILC MIC

v+

(+)

(-)

③

v-

PH3 IMD 5 min v-: 2 µm/hr v+: 3 µm/hr

30 µm With electric field

v+

(+)

MILC

With electric field

(c)

(d) Fig. 2



a-Si ① region

300 nm

300 nm

1 μm

② 110º

110º

MILC region

MILC growth direction

MILC region

cathode 1 μm


Over-etched MILC region

(a)

(b) 15 a-Si region

3001nm μm

③ MILC region


(c)

95º

anode

MILC rate (µm/hr)

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 35 36 37 38 39 40 41 42

Page 28 of 29

1 μm

10

anode without ɛ-field

5 cathode

0

0

1

2

3

4

Doping time (min) ACS Paragon Plus Environment

Fig. 3

(d)

5

Page 29 of 29


1 2 3 4 5 6 7 8 9 10 11 12MILC rate saturation by 13internal interfacial ɛ-field NiSi2 14 15 + 16 17 + 18 19 + Si 20 21 22 Enhancement of MILC 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42

c-Si

V

P P P

NiVNi+ by ɛ-field

MILC rate saturation by

a-Si

NiSi2

internal interfacial ɛ-field

c-Si VSi-

ɛ-field

a-Si

P+

Ni-

P+ P+

VNi+

Retardation of MILC by ɛ-field

PH3 doping

ɛ-field PH3 doping

(b)

(a)

Fig. 4 ACS Paragon Plus Environment

Electric-Field Enhanced Ni Induced Lateral Crystallization Rate

Recommend Documents