BSI Chapter 185 Joint Draft
Chapter 185 Joint Draft
Tuesday, December 19th.
Shenzhen, Su Chen's laboratory.
Su Chen spent the entire morning reviewing the fifth part written by Zhou Zhiyuan.
Part Five is the most sensitive part of the entire paper—the experimental data from the 300mm experiment are being published for the first time. Every number, every chart, and every description will be scrutinized by the reviewers with a magnifying glass.
He added a note to a description of an SEM cross-section on page seven:
"The scale bar in this figure needs to be re-labeled. The current scale bar shows 20 μm, but the actual magnification of the image should correspond to 10 μm. If reviewers find a mismatch between the scale bar and the image, they will directly question the reliability of the data."
He added another note in the error analysis table on page eleven:
"侧壁角度的测量误差±0.018°是三次独立测量的标準偏差。建议在表格下方加一行注释说明测量方法：'使用XXX型号扫描电镜，在晶圆中心、1/2半径、3/4半径三个位置各取5个测量点，共15个数据点，取标準偏差。'审稿人需要知道这个数字是怎幺来的。"
There are a total of eleven annotations.
At 2 PM, he sent the annotations to Zhou Zhiyuan: "Eleven annotations in Part Five. Focus on the scale on page seven and the error analysis on page eleven. The rest are minor wording adjustments."
Three hours later, Zhou Zhiyuan replied: "All revisions are complete. The scale has been re-marked, and the error analysis includes a description of the measurement methods. Please review it again."
Su Chen spent forty minutes checking the revised version.
"No problem. Part Five is finalized."
December 20th, Wednesday.
Su Chen began writing the final section of the sixth part of his thesis—"Future Directions".
This section only has three paragraphs, but he spent a whole day writing it.
The first paragraph discusses the theoretical predictions of the third-order modified model on larger-scale wafers. He includes the extrapolated data for 400mm here, labeled "theoretical extrapolation," along with the complete calculation parameters and boundary conditions.
The second paragraph discusses the potential applications of the third-order modified model in other thermoelastic coupling scenarios. It lists three directions: thin film stress control, thermal deformation compensation of nanoimprint dies, and temperature drift prediction of vibration sensors.
He paused for a long time when writing about the third direction.
Temperature drift of vibration sensors.
This is precisely Ishikawa Akira's field of research.
He did not mention Ishikawa Akira's name in the paper, nor did he cite Ishikawa Akira's papers. He only wrote a sentence at the end of this paragraph:
"The general framework of the third-order correction model theoretically makes it applicable to any micro/nano fabrication and sensing system dominated by thermoelastic coupling effects. Validating this universality requires independent experiments by researchers from different fields, and the authors of this paper look forward to such collaborations."
This statement is a public invitation—not to any one person, but to the entire academic community.
The third paragraph is the last paragraph of the entire text. Su Chen wrote:
"The third-order nonlinear extension proposed in this paper is not the end point of etching technology, but a theoretical starting point. A sidewall angle of 12.07° and a control accuracy of ±0.018° demonstrate the model's effectiveness on a 300mm scale, while the theoretical extrapolation to 400mm suggests further improvements on larger scales. These results collectively point to a fundamental fact: when we describe the physical processes of thermoelastic coupling with sufficiently precise mathematical tools, the engineering limits will be defined by physical laws rather than empirical formulas."
He read the passage three times. Then he changed "hinted at" to "foreshadowed". Then he changed "foreshadowed" back to "hinted at".
In the end, he kept the word "hinted".
Because the 400mm value has not yet been experimentally verified. Before experimental verification, theoretical extrapolation can only "hint" at it, not "predict".
The wording may be minor, but under Nature Materials' peer review standards, this difference can determine whether a reviewer requests a "minor revision" or a "major revision".
Thursday, December 21st.
Zhou Zhiyuan completed the unified compilation of all references in the text.
A total of 48 references were cited. Three were related to Bosch, all of which are indispensable classics in the field of DRIE. The 2019 thermal stress compensation paper that Su Chen had previously requested be removed was not included in the list.
He sent Su Chen a list of references: "48 articles. Please check if any need to be added or deleted."
Su Chen spent twenty minutes checking each item one by one.
"Delete paper number 31. It was a conference paper from 2017, and the data was later corrected by the authors themselves. Citations of outdated data will be picked up by reviewers. Replace it with a revised version published by the same authors in the Journal of Micromechanics and Microengineering in 2019."
Zhou Zhiyuan: "Okay. What about the rest?"
Su Chen: "Everything else is fine. 47 articles confirmed."
Friday, December 22nd.
3 PM.
Su Chen and Zhou Zhiyuan were online at the same time.
Su Chen sent the final draft of Part 6 of the paper to Zhou Zhiyuan. Zhou Zhiyuan then sent Su Chen the compiled references, acknowledgments, and supplementary materials.
Then the two of them started collaborating on the manuscript.
The collaborative drafting process was more complex than anticipated. The six sections were written by two different people. Although the notation system had been standardized beforehand, their writing styles still differed. Su Chen's section was more concise and direct, while Zhou Zhiyuan's was more detailed and focused on the readability of the derivation process.
They spent three hours adjusting the writing style of the entire text to make it as consistent as possible.
By 6 p.m., the combined manuscript was completed.
The full text consists of six parts, plus an abstract, introduction, references, and supplementary materials, totaling sixty-eight pages.
Title: Third-Order Nonlinear Extension of Thermoelastic Coupling Effect: Theoretical Framework, Experimental Verification, and Scale Extrapolation
Authors: Su Chen¹², Zhou Zhiyuan³
¹ Vilan Microsystems (Shenzhen) Co., Ltd.
² Department of Precision Instruments, Tsinghua University (PhD candidate)
³ Department of Precision Instruments, Tsinghua University
The last sentence of the abstract is:
"The third-order nonlinear extended framework proposed in this paper provides a general theoretical tool for the accurate description of thermoelastic coupling effects. Its applicability extends beyond specific etching processes and can be generalized to any micro/nano fabrication and sensing system dominated by thermoelastic coupling."
Zhou Zhiyuan looked at the completed paper after it was finalized and remained silent for a long time.
Then he typed: "Three months. From the first time you wrote down your idea for a third-order revision on the whiteboard in my office to the completion of the final draft, three months."
Su Chen replied, "It was because you rewrote Part Two from nine pages to seventeen pages that this paper qualified for submission to Nature Materials. If it had only been written according to JMEMS standards, the paper's impact would have been much smaller."
Zhou Zhiyuan: "The theory is yours. The framework is yours. The 300mm data is yours. I just repackaged these things in language that reviewers can understand."
Su Chen did not reply.
A little while later, he sent another message: "The submission deadline is the first week of January, after the New Year's Day holiday. I'll read through the entire article three more times in the next few days and make final revisions."
Zhou Zhiyuan: "Me too. We each read through it, and communicate immediately when we find any problems."
Su Chen: "Okay."
December 22nd, 8 PM.
Shenzhen, Lin Wei's office.
On the strategic draft in front of Lin Wei, a green checkmark has appeared next to the line "Late December: The paper is completed".
She picked up the phone and dialed Zhao Guoping: "The paper is finalized. Submit it in the first week of January."
Zhao Guoping: "Received. Regarding the product line progress—the contracts with the three new packaging companies have all been signed. Hongyuan Precision Equipment has confirmed delivery on January 8th. As for CR Microelectronics, last week's yield rate was 93.1%, ahead of schedule."
"Okay." Lin Wei added a few more annotations to the draft. "What about the alliance?"
"Four of the six companies slated to sign agreements have already done so. The remaining two are undergoing internal approval processes, which are expected to be completed in early January. At that time, the alliance will have 27 official members."
"What about a layered scheme?"
"The permissions for all eight core members have been activated. The technology sharing platform for standard members is currently in testing and is expected to officially launch in mid-January. MicroPort Sensing is being treated as a standard member, and there are currently no objections."
"Okay." Lin Wei closed her notebook. "Both the product line and the academic line are on track. The most important thing now is just one thing—paper peer review. We can't control that, but we can control ensuring that the submitted versions are flawless."
December 22nd, 9 PM.
Stuttgart, Germany.
Stein received a reply from Professor Albrecht in his email inbox.
The email was very short:
"Dr. Stein, I have carefully read all the analysis materials you compiled, especially the third folder, 'Inverse Derivation of the Third-Order Model.' Your conclusion is correct—without obtaining the complete derivation process, we cannot reconstruct Su Chen's third-order modified model. The mathematical tools he used go beyond the conventional thermoelastic coupling analysis framework and may involve some non-standard techniques in tensor algebra."
However, I noticed the direction you wrote on the whiteboard: 'material differentiation.' This idea is valuable. Bosch has thirty years of experience in silicon-based materials, which is something no startup can catch up to in a short time. If Su Chen's theory is indeed universally applicable, then the theory itself doesn't constitute a barrier—because everyone can use it after the paper is published. The real barrier lies in materials, equipment, and process experience.
I suggest you work with Professor Günther as soon as possible to draft a feasibility assessment for a material differentiation solution. I can provide technical support in the characterization of thermoelastic materials.
—Albrecht
After reading the email, Stein wrote a line in his notebook:
Albrecht confirmed: theory does not constitute a barrier; materials do. The direction is correct.
He then opened another folder and began drafting an outline for a material differentiation plan.
December 22nd, 10 PM.
Tokyo, Japan.
On Akira Ishikawa's computer screen, an Excel spreadsheet was densely packed with three years' worth of gyroscope temperature drift data.
He spent nearly two weeks organizing this data. Each set of data was labeled with the experimental conditions: temperature range, heating rate, holding time, and measurement frequency.
The data processing is complete.
But the analysis hasn't even started yet.
The analysis requires the complete mathematical derivation of Su Chen's third-order modified model. Without the complete derivation, he cannot fit the model into the parameter space of the gyroscope.
He closed Excel, opened his browser, and searched for Su Chen's recent activities.
No new papers. No new public speeches. Nothing at all.
"Writing a thesis."
Ishikawa Akira muttered to himself. He knew Su Chen must be writing a paper. He could tell from the tone of the email reply—"Let's communicate after the paper is published"—which meant the paper hadn't been submitted yet, but he was already writing it.
Based on Su Chen's work efficiency, the paper will likely be submitted around January. The review process takes six to eight weeks. In other words, he could see the complete theoretical framework as early as March.
March.
Three months of waiting.
Akira Ishikawa saved the data file and turned off the computer.
He walked to his desk, opened the first drawer, and glanced at the folded A3 sheet of paper. Two technical routes were drawn on the paper—Su Chen's DRIE and his gyroscope—connected by a dotted line with the words "thermoelastic coupling" written above them.
He closed the drawer.
Papers, etc.
December 22nd, 11 PM.
Shenzhen, Su Chen's dormitory.
Su Chen sat at his desk, his laptop in front of him displaying the completed paper after being edited together.
Page 68.
He started from the first page and read it word by word.
He paused when he reached the third section of the second part. He reread the sentence written by Zhou Zhiyuan twice: "The third-order nonlinear extension proposed in this section does not depend on the specific etching process or material system in its mathematical form, but is based on a general physical framework of thermoelastic coupling effect."
Three months ago, when he first wrote down his idea for third-order correction on the whiteboard in Zhou Zhiyuan's office, he was only thinking about solving the problem of sidewall angle control for 300mm DRIE.
It was Zhou Zhiyuan who transformed it into a general theoretical framework.
The experimental data from 300mm proves that this framework is not just mathematically elegant, but physically real.
The theoretical extrapolation of 400mm suggests the universality of this framework.
Now, these sixty-eight pages string all of this together into a complete story—from hypothesis to derivation, from derivation to verification, from verification to extrapolation, and from extrapolation to prospect.
Su Chen continued reading.
Outside the window, the Shenzhen night in late December was quiet and warm. A sliver of light from the distant city lights shone through the gaps in the curtains.
He read the last paragraph of the paper:
"...When we describe the physical processes of thermoelastic coupling with sufficiently precise mathematical tools, the engineering limits will be defined by physical laws rather than empirical formulas."
He did not change this sentence.
Then he turned back to the first page and read it a second time from the beginning.