Fullactivationofboroninsilicondopedbyself-assembledmolecularmonolayers资源-CSDN文库

63 浏览量 2021-02-11 21:58:05 上传评论收藏 1.68MB PDF 举报

### 全激活硼在硅中的自组装分子单层掺杂技术 #### 摘要与研究背景本文探讨了利用自组装分子单层（Self-Assembled Molecular Monolayer, SAMM）技术对硅进行硼掺杂的研究。该技术在纳米电子学领域展现出巨大的潜力，因其能够实现原子级精度的非破坏性掺杂，并适用于复杂的三维表面。然而，近期研究表明，SAMM技术引入的碳杂质显著降低了磷掺杂剂的激活率，导致形成多数载流子陷阱。因此，开发一种无缺陷、具有高激活率的SAMM掺杂技术对于可靠的应用至关重要。 #### 硼掺杂的全激活研究人员采用霍尔测量和二次离子质谱（Secondary Ion Mass Spectrometry, SIMS）技术来评估硼掺杂剂的激活率，并利用深能级瞬态光谱（Deep Level Transient Spectroscopy, DLTS）以及少数载流子瞬态光谱（Minority Carrier Transient Spectroscopy, MCTS）分析了通过SAMM技术在硅中掺入硼时产生的缺陷。研究发现，与磷掺杂剂不同的是，硼掺杂剂的激活率接近100%，这与缺陷测量结果（DLTS和MCTS）一致。仅有不到1%的硼掺杂剂与氧杂质结合形成多数空穴陷阱。值得注意的是，以CsH和CsOH形式存在的碳相关缺陷在硼掺杂的硅中作为少数载流子陷阱状态存在，这意味着它们只会捕获电子。因此，即使存在高浓度的碳杂质，也不会影响硼掺杂剂的激活率。 #### 关键词解释 - **全激活**：指掺杂剂完全转化为有效的载流子，即所有掺杂的原子都能够贡献于半导体的导电性能。 - **硼掺杂硅**：指在硅材料中掺入硼元素，以形成p型半导体。 - **分子单层掺杂**：通过在材料表面形成一层自组装的分子单层来进行掺杂的技术。 - **碳相关缺陷**：指由碳元素引起的缺陷，这些缺陷可能会影响半导体的性能。 - **少数载流子陷阱**：指能够捕获并暂时储存少数载流子的缺陷或陷阱状态。 #### 引言部分解析引言部分介绍了SAMM掺杂技术的特点及其在纳米电子学领域的应用前景。由于传统的掺杂技术可能导致晶格损伤，限制了其在复杂三维结构中的应用，SAMM技术提供了一种新的解决方案。通过在硅表面上自组装形成一层分子单层，可以实现精确控制的掺杂过程，这对于提高器件性能和可靠性至关重要。 #### 结论与展望本研究揭示了通过SAMM技术对硅进行硼掺杂时，即使存在碳杂质也不会显著影响硼掺杂剂的激活率。这一发现对于开发新型高效、可靠的半导体器件具有重要意义。未来的研究方向可能包括探索其他类型的掺杂剂和改进SAMM技术，以进一步提高器件性能并减少制造过程中的缺陷引入。此外，深入理解SAMM掺杂机制以及如何优化掺杂过程也是重要的研究课题。本文通过对硼掺杂硅的SAMM技术进行系统研究，为理解和改善新型半导体材料的性能提供了有价值的信息。随着该技术的不断进步，有望为纳米电子学领域带来革命性的变化。

资源推荐

资源详情

资源评论

Full Activation of Boron in Silicon Doped by Self-Assembled

Molecular Monolayers

Xuejiao Gao,

†,¶

Ilia Kolevatov,

‡

Kaixiang Chen,

†

Bin Guan,

†,#

Abdelmadjid Mesli,

Edouard Monakhov,

‡

and Yaping Dan*

,†

†

University of Michigan − Shanghai Jiao Tong University Joint Institute, Shanghai Jiao Tong University, Shanghai 200240, China

‡

Department of Physics, Center for Material Science and Nanotechnology, University of Oslo, Blindern, P.O. Box 1048, N-0316

Oslo, Norway

Institut Mate



riaux Microe



lectronique Nanosciences de Provence, UMR 6242 CNRS, Universite



Aix-Marseille, 13397 Marseille

Cedex 20, France

Supporting Information

ABSTRACT: Self-assembled molecular monolayer (SAMM) doping has

great potential in state-of-the-art nanoelectronics with unique features of

atomically precision and nondestructive doping on complex 3D surfaces.

However, it was recently found that carbon impurities introduced by the

SAMM signiﬁcantly reduced the activation rate of phosphorus dopants by

forming majority carrier traps. Developing a defect-free SAMM-doping

technique with a high activation rate for dopants becomes critical for

reliable applications. Considering that susbstitutional boron does not

interact with carbon in silicon, herein we employ Hall measurements and

secondary ion mass spectrometry (SIMS) to investigate the boron

activation rate and then deep level transient spectroscopy (DLTS) and

minority carrier transient spectroscopy (MCTS) to analyze defects in

boron-doped silicon by the SAMM technique. Unlike the phosphorus

dopants, the activation rate of boron dopants is close to 100%, which is

consistent with the defect measurement results (DLTS and MCTS). Only less than 1% boron dopants bind with oxygen

impurities, forming majority hole traps. Interestingly, carbon-related defects in the form of C

H and C

OH act as minority trap

states in boron-doped silicon, which will only capture electrons. As a result, the high concentration of carbon impurities has no

impact on the activation rate of boron dopants.

KEYWORDS: full activation, boron-doped silicon, molecular monolayer doping, carbon-related defects, minority carrier trap

■

INTRODUCTION

Self-assembled molecular monolayer (SAMM) doping enables

atomically precise and damage-free doping on 3D surfaces and

thus may ﬁnd important applications in developing deep

nanometer-scale electronics.

1−4

In this doping process, dopant-

containing molecular monolayers are ﬁrst covalently bonded

onto hydrogen-terminated semiconductor surfaces. A thermal

process is then employed to drive the dopants into the

semiconductor bulk, and the dopants are electrically activated

at the same time.

Ruess et al. ﬁrst demonstrated an ultrahigh

vacuum monolayer doping technique in 2004

and later

adopted scanning tunneling microscopy (STM) to covalently

graft single phosphine molecu les onto a silicon surface

followed by incorporating annealing to activate phosphorus.

7,8

For large-scale applications, a solution-based monolayer

doping was demonstrated

in 2008 where thermal energy

ultraviolet irradiation

was employed to generate surface

radicals for covalent attachment of doping molecules. Up to

now, this doping technology has been applied to sub-5 nm

ultrashallow junction formation,

conformal doping on ﬁn-

FETs,

and atomic-scale surface patterning.

12,13

In addition,

single dopants may be potentially controlled at large scale by

combining “top-down” large-scale nanofabrication and “bot-

tom-up” self-assembly.

One of the main concerns for solution-based SAMM doping

is organic contamination and dopant deactivation.

15−17

Shimizu et al.

have demonstrated that carbon and oxygen

diﬀused unintentionally into silicon during the the rmal

annealing process, and they suggested to sacriﬁce a top layer

of 2−3 nm for industrial application. Recently, we found that

carbon contaminants from organic molecules can electrically

deactivate at least 20% of the phosphorus.

15,18

Taheri et al.

proposed gas-phase molecule doping, which not only removed

the organic solvent but also increased the uniformity and

controllability.

Holmes and co-workers grew a thin SiO

layer to achieve a minimal carbon contamination.

However,

Received: November 8, 2019

Accepted: December 10, 2019

Published: December 10, 2019

Article

pubs.acs.org/acsaelm

Cite This: ACS Appl. Electron. Mater. XXXX, XXX, XXX−XXX

ACS Appl. Electron. Mater. XXXX, XXX, XXX−XXX

Downloaded via SHANGHAI JIAO TONG UNIV on January 2, 2020 at 09:48:47 (UTC).

See https://pubs.acs.org/sharingguidelines for options on how to legitimately share published articles.

all these investigations were on phosphorus-doped silicon.

Little attention was paid to how the carbon or oxygen in

organic carriers interact with boron dopants in silicon and

whether electrically active defects exist in boron-doped silicon

via SAMM.

In this work, we investigated boron-doped silicon by SAMM

from the aspects of activation rate and defects. Allylboronic

acid pinacol ester (ABAPE) molecules were applied to

introduce boron dopants into silicon by SAMM doping. The

hole and boron dopant concentrations were measured by Hall

eﬀect measurement and secondary ion mass spectrometry

(SIMS), respectively. To fully understand the defects in

SAMM-doped silicon, we employed deep leve l transient

spectroscopy (DLTS) and minority carrier transient spectros-

copy (MCTS) to investigate the majority and minority carrier

trap states, respectively. The results show that carbon

impurities do not bind with boron dopants. Carbon impurities

only form minority carrier trap states that have no impact on

the ionization rate of boron dopants, as a result of which boron

dopants are close to full activation. Only oxygen impurities will

bond with boron dopants, resulting in the deactivation of less

than 1% boron dopants.

■

RESULTS AND DISCUSSIONS

ABAPE monolayers were covalently bonded onto a silicon

surface following the process described in refs 1 and 15.A

compact and robust molecular monolayer was demonstrated

by co mparing X-ray photoelectron spectroscopic (XPS)

spectra of blank sample (freshly HF-etched Si), control sample

(ABAPE/p-xylene = 0:1), and ABAPE-functionalized sample

(ABAPE/p-xylene = 1:2), as shown in Figure 1 and

Supplementary Figures S2,S3.InFigure 1a, the silicon dioxide

peak (102.9 eV) is nearly invisible on the freshly HF-etched

silicon surface, and a Si 2p peak at 99.5 eV due to the Si

substrate splits into Si 2p (3/2) at 99.2 eV and Si 2p (1/2) at

99.8 eV. After heat treatment in p-xylene, the silicon dioxide

peak at 102.9 eV (fwhm 1.8 eV) shows up on the control

sample due to surface oxidation (Figure 1c). However, with a

compact ABAPE monolayer covering the silicon surface, only a

tiny silicon dioxide peak was detectable on the B-function-

alized silicon surface (Figure 1e), indicating that a compact

and robust monolayer of ABAPE has been successfully grafted.

No B 1s was detected on blank sample (Figure 1b) and control

sample (Figure 1d), but a broad peak of Si 2s plasma loss was

detected. In contrast, after functionalization with ABAPE, a

peak at 191.3 eV (fwhm 1.5 eV) was present in Figure 1f,

which is ascribed to the boron in the ABAPE monolayer. In

addition, C 1s narrow scans also support the successful

coupling of boron-containing molecules, as shown in

Supplementary Figure S3.

To evaluate the boron doping process, SAMM doping was

ﬁrst performed on a high-resistivity silicon wafer (>10 kΩ cm).

Experiment details can be found in our previous publication.

A control sample was prepared by removing ABAPE in p-

xylene solvent. van der Pauw four-probe measurements were

applied in darkness to measure the sheet resistance. Table 1

compares the experiment results of blank sample (bare silicon

wafer), control sample, and B-doped sample. The sheet

resistance of the control sample (299.60 kΩ/□) is very

close to that of blank sample (315.82 kΩ/□), which indicates

that SAMM doping introduces negligible contamination. After

boron was doped into silicon via SAMM, the sheet resistance

dramatically decreased to 3.20 kΩ/□. This drop in sheet

resistance is exclusively attributed to molecular monolayer

doping, and hence, eﬀective boron doping in silicon has been

demonstrated.

As we have discovered that carbon contamination from the

organic dopant carriers can deactivate as least 20% of the

phosphorus dopants in phosphorus-doped silicon by SAMM,

the investigations on the activation rate and defects in boron-

doped silicon became the keynote of this research. First, the

boron distribution near the silicon surface was analyzed by

SIMS. As shown in Figure 2a, the boron concentration rapidly

drops from 1.59 × 10

at the surface to 2.6 × 10

−3

at ∼5

nm below the surface. The boron dopants then drop to 6.1 ×

−3

at ∼104 nm below the surface. Because a monolayer

of ABAPE is grafted on the silicon surface, the total amount of

initial dopants is limited. The SIMS data from 5 to 104 nm

agrees well with the limited-source diﬀusion. By ﬁtting the

limited-source diﬀusion model into SIMS data,

adiﬀusivity of

7.24 × 10

−15

−1

was obtained, comparable to the

diﬀusivities for boron reported in the literature.

The sharp

decrease at the ﬁrst 5 nm was likely related with the observed

Figure 1. XPS narrow scans of blank sample, control sample, and

ABAPE-functionalized sample. (a) and (b) are Si 2p and B 1s narrow

scans on blank sample. (c) and (d) are Si 2p and B 1s narrow scans on

control sample. (e) and (f) are Si 2p and B 1s narrow scans on

ABAPE-functionalized sample. The takeoﬀ angle is 45°.

Table 1. Sheet Resistance of Blank Sample, Control Sample,

and Boron-Doped Samples by van der Pauw Technique

sheet resistance (kΩ/□)

blank sample 315.82

control sample 299.60

boron-doped sample 3.20

ACS Applied Electronic Materials Article

DOI: 10.1021/acsaelm.9b00748

ACS Appl. Electron. Mater. XXXX, XXX, XXX−XXX

剩余6页未读，继续阅读

评论收藏

内容反馈

weixin_38616033

粉丝: 2
资源: 931

Full activation of boron in silicon doped by self-assembled mole...

最新资源

Full activation of boron in silicon doped by self-assembled mole...

Dopant activation and photoresponses of boron-doped silicon by self-assembled molecular monolayers

Pattern formation by local self-activation and lateral inhibition

jakarta.activation-api-1.2.1-API文档-中文版.zip

jakarta.activation-api-1.2.2-API文档-中文版.zip

javax.activation-1.2.0-API文档-中文版.zip

jakarta.activation-api-1.2.2-API文档-中英对照版.zip

javax.activation-api-1.2.0-API文档-中文版.zip

jakarta.activation-1.2.2-API文档-中英对照版.zip

jakarta.activation-api-1.2.1-API文档-中英对照版.zip

activation.jar commons-email-1.2.jar mail.jar免费下载

activation-1.1.1.jar.zip

Multi-Resonant-Activation for a System Only Driven by Multi-State Noise

samsung-activation-bypass-v2跳过-45.zip

javax.activation-api-1.2.0-API文档-中英对照版.zip

C−H Bond Activation

keras-self-attention:处理顺序数据的注意力机制，考虑每个时间戳的上下文

Hebbian Learning and Negative Feedback Networks

activation-1.1.1-sources.jar

微软文库：Overview of volume activation of Office-35

基于CORDIC的反正弦和反余弦计算的FPGA实现

BA无标度网络中的SIR模型

使用3DCNN和卷积LSTM进行手势识别学习时空特征

基于三次贝塞尔曲线的类汽车曲率连续路径平滑

基于机器学习的设备剩余寿命预测方法综述

基于维纳过程的退化模型，具有递归过滤算法，可用于估计剩余使用寿命

基于FPGA的奇异值和特征值分解的快速实现。

磁悬浮系统自适应模糊PID控制器的设计

最新资源