Early stages of polycrystalline diamond deposition Laser reﬂectance at substrates with growing nanodiamonds

2025-05-03 0 0 1.78MB 32 页 10玖币

侵权投诉

Early stages of polycrystalline diamond deposition:

Laser reﬂectance at substrates with growing

nanodiamonds

David Vázquez-Cortés, Stoﬀel D. Janssens,

Burhannudin Sutisna, and Eliot Fried∗

Mechanics and Materials Unit (MMU), Okinawa Institute of Science and Technology

Graduate University (OIST), 1919-1 Tancha, Onna-son, Kunigami-gun, Okinawa, Japan

904-0495

Keywords: incubation period, polycrystalline diamond deposition, laser reﬂectance, seed

density, Rayleigh scattering, interferometry.

Abstract

The chemical vapor deposition of polycrystalline diamond (PCD) ﬁlms is typ-

ically done on substrates seeded with diamond nanoparticles. Specular laser re-

ﬂectance and a continuous ﬁlm model have been used to monitor the thickness of

these ﬁlms during their deposition. However, most seeds are isolated during the

early stages of the deposition, which questions the utility of applying such a con-

tinuous ﬁlm model for monitoring deposition before ﬁlm formation. In this work,

we present a model based on the Rayleigh theory of scattering for laser reﬂectance

at substrates with growing nanodiamonds to capture the early stages of PCD de-

position. The reﬂectance behavior predicted by our model diﬀers from that of a

continuous ﬁlm, which is well-described by the continuous ﬁlm model. This diﬀer-

ence enlarges as the seed density used in our model decreases. We verify this trend

∗Corresponding author: tel: +81(098)966-1372. E-mail: eliot.fried@oist.jp

arXiv:2210.00776v1 [cond-mat.mtrl-sci] 3 Oct 2022

experimentally by depositing diamond under identical conditions on substrates with

various seed densities. A relation derived from our model is used to ﬁt reﬂectance

data from which seed densities are obtained that are proportional to those found

with electron microscopy. We also show that relying on the continuous ﬁlm model

for describing the early stages of deposition can result in falsely deducing the exis-

tence of incubation, and that the continuous ﬁlm model can be used safely beyond

the early stages of deposition. Based on these ﬁndings, we delineate a robust method

for obtaining growth rates and incubation periods from reﬂectance measurements.

This work may also advance the general understanding of nanoparticle growth and

formation.

1 Introduction

1.1 Motivation and goals

According to Russel [1], the term “incubation time”, which is used interchangeably with

“incubation period” [2], “induction time” [3], and “induction period” [4], describes ther-

mally activated processes that do not commence immediately after establishing a reaction

temperature. A prime example of a thermally activated process is the chemical vapor

deposition (CVD) of polycrystalline diamond (PCD) [5]. Due to high activation energies

associated with the nucleation of diamond grains on foreign substrates, relative to those

of diamond deposition, incubation periods preceding deposition are observed [6–8]. Such

incubation periods can be reduced signiﬁcantly, or even eliminated entirely, by seeding

diamond nanoparticles, typically detonation nanodiamonds, on substrates before depo-

sition [9]. Still, the seeding step does not prevent incubation under certain deposition

conditions [8,10,11]. Incubation can be related to the etching of seeds [2], which aﬀects the

seed density. Consequently, the morphological features of ﬁlms and the physical properties

such as thermal conductivity, transparency, and adhesiveness are also aﬀected [12–14]. A

fundamental understanding of incubation is therefore essential for producing PCD ﬁlms

with tailored properties. The occurrence of incubation is mainly deduced from monitor-

ing deposition rates during the early stages of deposition using specular laser reﬂectance

interferometry. This measurement method is typically based on a model that assumes

continuous ﬁlm interference [2, 8]. However, seeds are isolated during the early stages

of deposition, raising questions about whether such a model can be reliably applied to

processes before ﬁlm formation. The goals of this work are to:

•Develop a semiquantitative, non-interference, and non-continuous ﬁlm model for

specular laser reﬂectance based on nanodiamond particle scattering that captures

the early stages of PCD CVD on seeded substrates.

•Test our model experimentally by depositing diamond on substrates with diﬀerent

seed densities and analyzing the samples with several microscopy techniques.

•Develop a method to estimate incubation periods, which we identify with a delay in

deposition and/or a reduced deposition rate during the early stages of deposition,

with specular laser reﬂectance.

For brevity, “specular laser reﬂectance” is shortened to “laser reﬂectance”. The term

“diamond particle” refers to a single grain or to a cluster of grains isolated from other

diamond particles and the term “seed” refers to a diamond particle deposited on the

surface of a substrate before deposition.

1.2 Background

The small lattice parameter and the high surface energies of diamond limit diamond

heteroepitaxy to diamond-on-iridium [15–20]. An alternative method to deposit diamond

on foreign substrates relies on seeding the substrate surface with diamond particles [21–

26]. Most work on seeding aims to increase the areal (seed) density with the objective of

minimizing the thickness at which ﬁlms become pinhole-free [27–29]. However, high seed

densities are not always desirable. For example, Mandal et al. [30] found that a low seed

density reduces stress in a PCD ﬁlm grown on an aluminum nitride substrate, aﬀording

the deposition of a thicker ﬁlm without delamination. Tsigkourakos et al. [31] found that

lowering the seed density increases the electrical conductance of boron-doped ﬁlms by

0 4 8 12 16

laser

photodetector

sample

Iin

Iout =α1α2α3Iin

diamond grains

ν2D

reﬂectance

time (a.u.)

α1Iin α1α2Iin

CVD system

silicon

a) b)

Figure 1: Schematic of a reﬂectance setup. a) Schematic of a setup for measuring

the specular laser reﬂectance of a sample in a chemical vapor deposition (CVD) system

during deposition. A laser beam of intensity Iin enters the CVD system at an angle

θ. After interaction with the windows and the sample, the beam exits the system with

intensity Iout. The arrows indicate the propagation direction of the laser beam. The

windows cause attenuations α1and α3, and attenuation α2is caused by the sample. The

inset of a) shows a schematic of diamond particles during the outset of deposition, with

seed density ν2D. b) A typical graph obtained by plotting the photodetector’s signal in a)

as a function of time. Regular oscillations, which are used for deposition rate estimation,

are observed for thin-ﬁlm interference.

one order of magnitude, and Janssens et al. [32] showed, through simulations, that the

seed density strongly aﬀects the grain size distribution. These results show that the seed

density is a useful parameter for tuning the properties of PCD ﬁlms and is therefore

explored systematically in this work.

In-situ laser reﬂectance interferometry is widely used for monitoring the thickness of a

nanocrystalline diamond ﬁlm during deposition. This is done by shining a laser beam on

a sample, collecting the specularly reﬂected light with a photodetector, and analyzing the

output signal that the photodetector produces [33–35]. Figure 1 shows a schematic of a

reﬂectance setup installed on a CVD reactor. The ﬁlm thickness and deposition rate are

typically calculated from the extrema of the oscillating reﬂectance and the time intervals

separating these extrema, respectively [4,36,37]. The oscillations are caused by thin ﬁlm

interference.

In the early stage of diamond deposition on seeded silicon substrates, laser reﬂectance

decays more slowly than expected for a continuous ﬁlm [18, 37]. Some authors have

related the initial slow decay of the reﬂectance to a reduced deposition rate or to the

varying roughness of the sample surface [2,36]. For the deposition of diamond on seeded

substrates, the existence of incubation is mainly derived from the slow decay in reﬂectance

during the early stages of deposition. However, it is not clear whether the estimation

of an incubation period is strongly aﬀected by the assumption of a continuous ﬁlm.

Here, we investigate this systematically by comparing laser reﬂectance with atomic force

microscopy (AFM) and scanning electron microscopy (SEM).

Bonnot et al. [38] showed that during the early stages of PCD deposition, scattering

from isolated diamond seeds increases with time. This increase was measured by in-

stalling a detector away from the specularly reﬂected laser. The analysis of the measured

scattering was done using a model based on Rayleigh scattering, assuming that scattering

caused by seeds is proportional to the square of their volume. This made it possible to

calculate the diamond deposition rate using scattering instead of reﬂectance. As demon-

strated by Smolin et al. [39], Mie scattering, which generalizes Rayleigh scattering to large

particles, can also explain the reﬂectance behavior during the initial stage of deposition.

Smolin et al. measured reﬂectance by introducing a laser beam through a quartz window

directed normal to the substrate surface and by measuring the specularly reﬂected beam

intensity by a photodiode. However, the experiments reported in their work were limited

by the control over seeding available at the time. Also, the implications of varying the

diamond seed density on reﬂectance were not systematically investigated. These works

demonstrate that light scattering, rather than interference, can accurately describe light–

sample interactions during the early stages of PCD deposition. However, to the best of

our knowledge, a relation that describes the reﬂectance during the early stages of PCD

deposition has not yet appeared in the literature. Moreover, we are not aware of any pre-

vious attempt to include light scattering phenomena in the analysis of reﬂectance curves

measured during incubation. We, therefore, develop an elementary model based on the

Rayleigh theory of scattering to explain the behavior of laser reﬂectance during the early

stages of PCD deposition and test the model experimentally. Due to the non-interference

nature of the light–sample interaction in the early stages of PCD deposition, we avoid

文档加载中……请稍候！
如果长时间未打开，您也可以点击刷新试试。

下载文档到电脑，查找使用更方便

10 玖币 0人已下载

立即下载

摘要：

Earlystagesofpolycrystallinediamonddeposition:LaserreectanceatsubstrateswithgrowingnanodiamondsDavidVázquez-Cortés,StoelD.Janssens,BurhannudinSutisna,andEliotFried*MechanicsandMaterialsUnit(MMU),OkinawaInstituteofScienceandTechnologyGraduateUniversity(OIST),1919-1Tancha,Onna-son,Kunigami-gun,Okina...

展开>> 收起<<

Early stages of polycrystalline diamond deposition Laser reﬂectance at substrates with growing nanodiamonds.pdf

共32页,预览5页

还剩页未读，继续阅读

声明：本站为文档C2C交易模式，即用户上传的文档直接被用户下载，本站只是中间服务平台，本站所有文档下载所得的收益归上传人(含作者)所有。玖贝云文库仅提供信息存储空间，仅对用户上传内容的表现方式做保护处理，对上载内容本身不做任何修改或编辑。若文档所含内容侵犯了您的版权或隐私，请立即通知玖贝云文库，我们立即给予删除！

Early stages of polycrystalline diamond deposition Laser reﬂectance at substrates with growing nanodiamonds

相关推荐

开通VIP享超值会员特权

作者详情

相关内容

热门标签

举报选择: