Chinese semiconductor industry

Status
Not open for further replies.

PopularScience

Junior Member
Registered Member
Summary
1. Planar Grating Laser Interferometer has been delayed by 2 years.
2. Fortunately, Harbin Institute of Technology got it done at the end of last year.
3. Now 28nm Lithography double exposure supports 14nm process。

消息人士透露,虽然比原计划推迟了2年,但是目前进展已经顺利。
大家都知道,所谓的0贰砖项的GKJ,就是28nm制程节点的GKJ。按照项目目标规定,该GKJ只能用于28nm节点,无法通过多重曝光应用于14nm节点。
虽然都是采用duv光源的浸润式原理的机器,但是14nm比28nm的机器的各项精度都要提升,其中最重要的技术难点就是GKJ内部测量的定位精度要大幅提高,因此14nmGKJ的内部测量仪器就必须从28nmGKJ的多轴激光干涉仪,升级为平面光栅激光干涉仪,只有这样才能满足掩模工件台、硅片双工件台和投影物镜之间复杂的相对位置、姿态测量需求,保证光刻机的整体套刻精度。
2023年之前,这个平面光栅激光干涉仪没有搞定,所以GKJ也就是28nm节点适用,无法更进一步提升至14nm,要等下一代改进版才行,不过,幸运的是哈工大这块去年底搞定了,原本全球GKJ巨头们都必须依赖英国和美国的平面光栅激光干涉仪,但是现在中国的光刻机不再需要了。
消息人士透露:“哈工大的这个激光干涉仪已应用在350nm至28nm的所有国产GKJ上了。”
新鲜出炉的平面光栅(极其昂贵)图:
华卓原本计划是用多轴激光干涉仪来实现符合28nm精度的双工件台,因为ASML在28nm节点的机器就是如此,到了14nm节点的时候ASML才转换到了平面光栅,可是华卓没有如同ASML那样顺利搞定,而是卡在这里相当长一段时间,华卓感觉难度很大,虽然也能最终搞定,但是必须比原计划大大的推迟了。 有了这个平面光栅,就解决了华卓的燃眉之急。 顺便也提升了双工件台和整个机器的精度,“使得目前这个28nm节点浸润式DUV光KJ,可以通过双重曝光支持14nm制程”——消息人士透露。
 

tonyget

Senior Member
Registered Member
smartfortwo:哈工大激光干涉仪分辨率77皮米,华卓精科光栅分辨率50皮米,差距不是很大,主要还是你说的可以提升产率。
smartfortwo:我看华卓精科招股书dws-i本来计划用的是平面光栅,难道是等不及先上了激光干涉仪?
havok:华卓也有自己研发的激光干涉仪,肯定是谁更好用谁的

smartfortwo:The laser interferometer of Harbin Institute of Technology has a resolution of 77 picometers, and the Huazhuo Jingke grating has a resolution of 50 picometers. The gap is not very big, mainly due to what you said can increase productivity.
smartfortwo:According to the prospectus of Huazhuo Jingke, dws-i originally planned to use a flat grating. Could it be that the laser interferometer can't be waited for?
havok:Huazhuo also has a laser interferometer developed by itself, it must be better to use whose
 

european_guy

Junior Member
Registered Member
Summary
1. Planar Grating Laser Interferometer has been delayed by 2 years.
2. Fortunately, Harbin Institute of Technology got it done at the end of last year.
3. Now 28nm Lithography double exposure supports 14nm process。

消息人士透露,虽然比原计划推迟了2年,但是目前进展已经顺利。
大家都知道,所谓的0贰砖项的GKJ,就是28nm制程节点的GKJ。按照项目目标规定,该GKJ只能用于28nm节点,无法通过多重曝光应用于14nm节点。
虽然都是采用duv光源的浸润式原理的机器,但是14nm比28nm的机器的各项精度都要提升,其中最重要的技术难点就是GKJ内部测量的定位精度要大幅提高,因此14nmGKJ的内部测量仪器就必须从28nmGKJ的多轴激光干涉仪,升级为平面光栅激光干涉仪,只有这样才能满足掩模工件台、硅片双工件台和投影物镜之间复杂的相对位置、姿态测量需求,保证光刻机的整体套刻精度。
2023年之前,这个平面光栅激光干涉仪没有搞定,所以GKJ也就是28nm节点适用,无法更进一步提升至14nm,要等下一代改进版才行,不过,幸运的是哈工大这块去年底搞定了,原本全球GKJ巨头们都必须依赖英国和美国的平面光栅激光干涉仪,但是现在中国的光刻机不再需要了。
消息人士透露:“哈工大的这个激光干涉仪已应用在350nm至28nm的所有国产GKJ上了。”
新鲜出炉的平面光栅(极其昂贵)图:
华卓原本计划是用多轴激光干涉仪来实现符合28nm精度的双工件台,因为ASML在28nm节点的机器就是如此,到了14nm节点的时候ASML才转换到了平面光栅,可是华卓没有如同ASML那样顺利搞定,而是卡在这里相当长一段时间,华卓感觉难度很大,虽然也能最终搞定,但是必须比原计划大大的推迟了。 有了这个平面光栅,就解决了华卓的燃眉之急。 顺便也提升了双工件台和整个机器的精度,“使得目前这个28nm节点浸润式DUV光KJ,可以通过双重曝光支持14nm制程”——消息人士透露。


I found this
Please, Log in or Register to view URLs content!
that seems to confirm that Harbin Institute of Technology built a prototype Grating Interferometer not later than September 2022 (paper's date).
 

PopularScience

Junior Member
Registered Member
Summary
1. Planar Grating Laser Interferometer has been delayed by 2 years.
2. Fortunately, Harbin Institute of Technology got it done at the end of last year.
3. Now 28nm Lithography double exposure supports 14nm process。

消息人士透露,虽然比原计划推迟了2年,但是目前进展已经顺利。
大家都知道,所谓的0贰砖项的GKJ,就是28nm制程节点的GKJ。按照项目目标规定,该GKJ只能用于28nm节点,无法通过多重曝光应用于14nm节点。
虽然都是采用duv光源的浸润式原理的机器,但是14nm比28nm的机器的各项精度都要提升,其中最重要的技术难点就是GKJ内部测量的定位精度要大幅提高,因此14nmGKJ的内部测量仪器就必须从28nmGKJ的多轴激光干涉仪,升级为平面光栅激光干涉仪,只有这样才能满足掩模工件台、硅片双工件台和投影物镜之间复杂的相对位置、姿态测量需求,保证光刻机的整体套刻精度。
2023年之前,这个平面光栅激光干涉仪没有搞定,所以GKJ也就是28nm节点适用,无法更进一步提升至14nm,要等下一代改进版才行,不过,幸运的是哈工大这块去年底搞定了,原本全球GKJ巨头们都必须依赖英国和美国的平面光栅激光干涉仪,但是现在中国的光刻机不再需要了。
消息人士透露:“哈工大的这个激光干涉仪已应用在350nm至28nm的所有国产GKJ上了。”
新鲜出炉的平面光栅(极其昂贵)图:
华卓原本计划是用多轴激光干涉仪来实现符合28nm精度的双工件台,因为ASML在28nm节点的机器就是如此,到了14nm节点的时候ASML才转换到了平面光栅,可是华卓没有如同ASML那样顺利搞定,而是卡在这里相当长一段时间,华卓感觉难度很大,虽然也能最终搞定,但是必须比原计划大大的推迟了。 有了这个平面光栅,就解决了华卓的燃眉之急。 顺便也提升了双工件台和整个机器的精度,“使得目前这个28nm节点浸润式DUV光KJ,可以通过双重曝光支持14nm制程”——消息人士透露。
I found this
Please, Log in or Register to view URLs content!
that seems to confirm that Harbin Institute of Technology built a prototype Grating Interferometer not later than September 2022 (paper's date).
Correct. Tan Jiu Bin is the academician of Chinese Academy of Science.
 

european_guy

Junior Member
Registered Member
I'd would like to highlight one technical aspect that was not so clear to me. So I write here hoping it will be useful.

We know that ASML lithography DUV immersion machines can be used for nodes from 40nm down to even 7nm. But what is the difference between all these machines?

Actually all these ASML machines have the same optical of 1.35NA. It means the optical system that goes for 28nm goes also for 7nm!

Now it seems confirmed that also SMEE immersion machine has a 1.35NA, because this corresponds to 38nm resolution, that is the same of ASML. Even the newest
Please, Log in or Register to view URLs content!
has 38nm resolution with 1.35NA.

So what is the difference among these machines (apart from productivity)? It seems a critical spec is the MMO (matched machine overlay), i.e the precision of the positioning of the wafer stage. Here below we can see the road map of ASML machines:


ASML_road_map.png


With this information clarified, I hope we can better follow all the ongoing discussions on interferometers with and without gratings etc.

So, because all ASML DUVi machines and also the SMEE's one, have Arf 193nm immersion 1.35NA, is the precision of the wafer stage that sets the limit of the node that is possible to produce.

--------------------------------------------------------------------------------------------------------------------------------

Considering the above, if ASML will be forced to ban DUVi, then they will ban the entire immersion line (but in this case they have to say goodby to the 30 and plus machines just ordered by SMIC with delivery 2024), or they will ship anyhow the machines but with reduced MMO, so to limit them for advanced nodes. The latter option makes more sense both technically and economically.
 
Last edited:

tonyget

Senior Member
Registered Member
So what is the difference among these machines (apart from productivity)? It seems a critical spec is the MMO (matched machine overlay), i.e the precision of the positioning of the wafer stage. Here below we can see the road map of ASML machines:


With this information clarified, I hope we can better follow all the ongoing discussions on interferometers with and without gratings etc.

So, because all ASML DUVi machines and also the SMEE's one, have Arf 193nm immersion 1.35NA, is the precision of the wafer stage that sets the limit of the node that is possible to produce.

havok already explained it,it's the machine overlay which involves interferometer determines the accuracy
 

WTAN

Junior Member
Registered Member
Cross post from the other thread, @WTAN Sir your comment it's either 70 or 90 that can support an EUVL machine.

Bro should be posted in the Semiconductor thread, what interest me is this " Started in 2019 and expected to be completed by the end of 2025, the HEPS can provide 60 to 70 beamlines at high brightness with more than 90 experimental stations."

The HEPS Synchrotron built by the CAS is certainly testing prototype equipment and doing R&D for a future SSMB EUVL Light Source Facility.
This future facility could be the planned Tsinghua Synchrotron or another CAS built and operated SSMB EUVL Facility.
I mentioned a while ago that the Older Shanghai Synchrotron was doing SSMB Research and this was confirmed recently in a post about the Tsinghua Uni SSMB Project.
This proves that the HEPS will be also constructed in such a way to allow for SSMB Research to be carried out.
China has a history of building 100% local projects in parallel with the foreign invested projects.
In this case the CAS will likely build a SSMB EUVL Facility in parallel with the Tsinghua University SSMB EUVL Facility.
 

WTAN

Junior Member
Registered Member

Han's Laser: Proximity lithography machine has been put into the market, stepping lithography machine has started user optimization​


Han’s Laser said in an institutional survey that at present, the company’s lithography machine project is mainly used in the field of discrete devices, with a resolution of 3-5 μm; among them, the proximity lithography machine has been put into use. In the market, the stepper lithography machine has started user optimization. The company will continue to formulate R&D plans according to market demand and business development, and increase the localization of R&D core components.

Please, Log in or Register to view URLs content!

Interesting exposition in Hubei.
View attachment 106746
View attachment 106745
Kind of a pulsed CO2 laser seed for a EUV light sources along with molten tin plasma droplet generators, nano-metrology tools and "microenvironment control tools".
Good find Tokenanalyst.
This looks like a first glimpse of the Subsystems for a Chinese made LPP EUVL Machine.
For them to display these products at a Exposition in Hubei would mean they are matured products already used in a prototype EUVL Machine.
 
Status
Not open for further replies.
Top