打开APP
userphoto
未登录

开通VIP,畅享免费电子书等14项超值服

开通VIP
Tech Brief: A Look at Atomic Layer Deposition (ALD...

转自:Lam Research

ALD(atomic layer deposition)在先进制程中使用得越来越广泛。与传统CVD相比,ALD沉积的薄膜均匀地覆盖在衬底表面结构上,具有极佳的conformality。ALD在side-wall imaging process(即SADP、SAQP)中是不可缺少的,更多新的应用也在研发之中。这篇文章对ALD技术做了一个比较简要的介绍,附在这里供同仁参考。

以下是原文:

Imagine being able to deposit a film of material just a few atomic layers at a time. As impossible as that sounds, atomic layer deposition (ALD) is a reality. In fact, it’s being used in an ever-increasing number of applications as an extremely precise and controllable process for creating thin films. Together with its etch counterpart –  – ALD is enabling the use of new materials and three-dimensional designs in advanced chip manufacturing. In this Tech Brief, we’ll take a closer look at this important atomic-scale process.

Chemical Vapor Deposition

ALD is actually a type of chemical vapor deposition (CVD), one of the most common methods of producing thin films during chip-making. In CVD, gaseous “precursor” chemicals flow into a process chamber that contains the silicon wafer. These precursors react on the wafer surface, forming the desired film along with byproducts that are removed from the chamber. Plasma-enhanced CVD (PECVD) uses a plasma to lower the deposition temperature while maintaining good film quality and high deposition rates. This is an important advantage since some CVD processes involve higher temperatures than can be tolerated by materials used in today’s advanced semiconductors. However, the PECVD process is “soupy” – precursors, plasma, byproducts, and other molecular fragments and species are all floating around in the chamber, making it difficult to control film deposition to the atomic scale.

Atomic Layer Deposition

The secret to gaining improved control was to split the deposition process into half-reactions, each of which can be well-controlled. The ALD process starts by flooding the reaction chamber with a precursor that coats (or “adsorbs” onto) the exposed surface of the wafer. This process is called self-limiting because the precursor can only adsorb onto exposed areas; once all those are covered, the adsorption stops. A second gas is then introduced and reacts with the precursor to form the desired material. This second step is also self-limiting:  once the available precursor sites are used up, the reaction stops. The two steps are repeated until the desired film thickness is obtained.

There are a couple of ways to divide the steps. In one technique known as spatial ALD, the wafer is moved between different locations and exposed to a different precursor at each. Another approach is to hold the wafer in one place and alternately introduce/remove precursors into the chamber. Known as temporal ALD, this method enables the wafer to be processed in a more symmetric environment, improving process results such as better critical dimension (CD) range control.

ALD Advantages

ALD offers a number of advantages, all of which arise from the self-limiting, sequential reactions. First, while deposition is not exactly a single atomic layer per cycle, film thickness is well controlled and excellent uniformity can be achieved across the wafer. Perhaps even more importantly, ALD creates layers that conform extremely well to the wafer topography, with identical film thicknesses deposited on the tops, sides, and bottoms of device features. This high conformality is a critical capability for high-aspect-ratio and 3D structures. Lastly, surfaces created by ALD are atomically smooth, with well-controlled chemical composition.

 

Applications

The ALD process is capable of creating both dielectric (insulating) and metal (conducting) films, depending on the choice of precursors. Its many advantages have led ALD to be used in several applications, and here we take a brief look at some of these.

Self-Aligned Patterning:  ALD plays a key role in self-aligned , which is used to form patterns smaller than can be produced with current lithography technology. In this technique, a thin spacer is deposited on pre-defined features. This spacer film must be highly conformal and very uniform as it will eventually define the critical dimensions of the final pattern.

3D NAND:  The three-dimensional structures in  memory devices require a high degree of process variability control. ALD is well suited for this and is used to form dielectric films on the sidewalls of memory holes. Metal ALD is also used for the word line fill in replacement-gate schemes, which require lateral deposition that completely fills narrow, horizontal features.

FinFET:  The thin gate sidewall spacers in  must be formed with extremely uniform thickness and no pinholes. ALD is an excellent way of depositing this layer, which separates the control gate from the three-dimensional fin structure.

 

Future Applications

The uses for ALD continue to grow. For example, one promising application is selective area deposition, which leverages inherently selective films. Researchers are currently developing ways to deposit metals and dielectrics in very specific locations – essentially creating a different patterning method. For the first time, selectivity is the most important film property and will be essential for integration at the 5 nm to 3 nm technology nodes. ALD is also being explored as a means to improve overlay control, or how precisely a new pattern can be aligned over an existing pattern. Any offset or misalignment to underlying electrical contacts can reduce conduction and negatively impact chip performance.

Looking ahead, we expect atomic layer processing to play an increasingly important role in advancing semiconductor manufacturing as these and other applications develop. Proven as a key enabling technology, ALD continues to evolve for use in challenging new structures and scaling strategies as they are integrated into next-generation devices.

注:本文转自Lam Research感谢原作者的付出,如果侵犯了您的相关权益,请联系后台,我们会及时处理!

本站仅提供存储服务,所有内容均由用户发布,如发现有害或侵权内容,请点击举报
打开APP,阅读全文并永久保存 查看更多类似文章
猜你喜欢
类似文章
Silicon alternatives key to future computers, consumer electronics
Platinum
Tesla Silicon Anode for 4680 Battery Cell: What's ...
Regular expressions • stringr
U Contrast Masking
Online handwriting recognition using multi convolution neural networks
更多类似文章 >>
生活服务
热点新闻
分享 收藏 导长图 关注 下载文章
绑定账号成功
后续可登录账号畅享VIP特权!
如果VIP功能使用有故障,
可点击这里联系客服!

联系客服