align:start position:0% the following content is provided under align:start position:0% the following content is provided under align:start position:0% the following content is provided under a creative commons license your support align:start position:0% a creative commons license your support align:start position:0% a creative commons license your support will help mit opencourseware continue to align:start position:0% will help mit opencourseware continue to align:start position:0% will help mit opencourseware continue to offer high quality educational resources align:start position:0% offer high quality educational resources align:start position:0% offer high quality educational resources for free align:start position:0% for free align:start position:0% for free to make a donation or to view additional align:start position:0% to make a donation or to view additional align:start position:0% to make a donation or to view additional materials from hundreds of mit courses align:start position:0% materials from hundreds of mit courses align:start position:0% materials from hundreds of mit courses visit mit opencourseware at ocw.mit.edu align:start position:0% align:start position:0% good morning everyone align:start position:0% good morning everyone align:start position:0% good morning everyone good evening in singapore align:start position:0% good evening in singapore align:start position:0% good evening in singapore where is everyone align:start position:0% align:start position:0% i think they're on the way we'll be here align:start position:0% i think they're on the way we'll be here align:start position:0% i think they're on the way we'll be here in a minute okay align:start position:0% in a minute okay align:start position:0% in a minute okay i'm gonna align:start position:0% i'm gonna align:start position:0% i'm gonna go ahead and align:start position:0% go ahead and align:start position:0% go ahead and uh start what i wanted to do is uh i align:start position:0% uh start what i wanted to do is uh i align:start position:0% uh start what i wanted to do is uh i sent a message around about these team align:start position:0% sent a message around about these team align:start position:0% sent a message around about these team guidelines so align:start position:0% guidelines so align:start position:0% guidelines so our team project guidelines so i'm not align:start position:0% our team project guidelines so i'm not align:start position:0% our team project guidelines so i'm not going to go in great detail through align:start position:0% going to go in great detail through align:start position:0% going to go in great detail through these slides hopefully everybody is well align:start position:0% these slides hopefully everybody is well align:start position:0% these slides hopefully everybody is well on the way towards identifying a team align:start position:0% on the way towards identifying a team align:start position:0% on the way towards identifying a team but i wanted to go quickly through this align:start position:0% but i wanted to go quickly through this align:start position:0% but i wanted to go quickly through this and then answer any questions you may align:start position:0% and then answer any questions you may align:start position:0% and then answer any questions you may have on the on the team projects align:start position:0% have on the on the team projects align:start position:0% have on the on the team projects a quick reminder that the team projects align:start position:0% a quick reminder that the team projects align:start position:0% a quick reminder that the team projects are about 20 align:start position:0% are about 20 align:start position:0% are about 20 uh of the grade so they are an important align:start position:0% uh of the grade so they are an important align:start position:0% uh of the grade so they are an important part of this align:start position:0% part of this align:start position:0% part of this and the basic expectation is this is align:start position:0% and the basic expectation is this is align:start position:0% and the basic expectation is this is where you get a chance to dive deep into align:start position:0% where you get a chance to dive deep into align:start position:0% where you get a chance to dive deep into the analysis align:start position:0% the analysis align:start position:0% the analysis and exercising and some of the kinds of align:start position:0% and exercising and some of the kinds of align:start position:0% and exercising and some of the kinds of tools and approaches we've been talking align:start position:0% tools and approaches we've been talking align:start position:0% tools and approaches we've been talking about with a set of data align:start position:0% about with a set of data align:start position:0% about with a set of data our hope is that you can find data align:start position:0% our hope is that you can find data align:start position:0% our hope is that you can find data yourself align:start position:0% align:start position:0% through the various members of the class align:start position:0% through the various members of the class align:start position:0% through the various members of the class i think align:start position:0% i think align:start position:0% i think in one of the earlier problem sets we align:start position:0% in one of the earlier problem sets we align:start position:0% in one of the earlier problem sets we asked for align:start position:0% asked for align:start position:0% asked for your experiences perhaps with some past align:start position:0% your experiences perhaps with some past align:start position:0% your experiences perhaps with some past data i think there's plenty of data align:start position:0% data i think there's plenty of data align:start position:0% data i think there's plenty of data around align:start position:0% around align:start position:0% around if you are not able to tap into some align:start position:0% if you are not able to tap into some align:start position:0% if you are not able to tap into some past data you you or colleagues have align:start position:0% past data you you or colleagues have align:start position:0% past data you you or colleagues have generated in the lab in a previous job align:start position:0% generated in the lab in a previous job align:start position:0% generated in the lab in a previous job that you can share align:start position:0% that you can share align:start position:0% that you can share or whatever you can also tap into the align:start position:0% or whatever you can also tap into the align:start position:0% or whatever you can also tap into the literature align:start position:0% literature align:start position:0% literature so if you find data in the literature align:start position:0% so if you find data in the literature align:start position:0% so if you find data in the literature that is perfectly fine as well align:start position:0% that is perfectly fine as well align:start position:0% that is perfectly fine as well um and then in a few cases we actually align:start position:0% um and then in a few cases we actually align:start position:0% um and then in a few cases we actually have some data from our own align:start position:0% have some data from our own align:start position:0% have some data from our own experimentation align:start position:0% experimentation align:start position:0% experimentation and research so if you're really stuck align:start position:0% and research so if you're really stuck align:start position:0% and research so if you're really stuck uh let me know or let hayden know and we align:start position:0% uh let me know or let hayden know and we align:start position:0% uh let me know or let hayden know and we can try to connect you up to something align:start position:0% align:start position:0% i've listed a few topic areas i don't align:start position:0% i've listed a few topic areas i don't align:start position:0% i've listed a few topic areas i don't want these to be align:start position:0% want these to be align:start position:0% want these to be limiting but i mean these are the main align:start position:0% limiting but i mean these are the main align:start position:0% limiting but i mean these are the main main kinds of topics we've been dealing align:start position:0% main kinds of topics we've been dealing align:start position:0% main kinds of topics we've been dealing with in the class things like process align:start position:0% with in the class things like process align:start position:0% with in the class things like process diagnosis looking at the data to detect align:start position:0% diagnosis looking at the data to detect align:start position:0% diagnosis looking at the data to detect uh problems align:start position:0% uh problems align:start position:0% uh problems ways of optimizing or improving the align:start position:0% ways of optimizing or improving the align:start position:0% ways of optimizing or improving the process through align:start position:0% process through align:start position:0% process through design of experiments optimization and align:start position:0% design of experiments optimization and align:start position:0% design of experiments optimization and so on and then some other sort of align:start position:0% so on and then some other sort of align:start position:0% so on and then some other sort of advanced applications things like align:start position:0% advanced applications things like align:start position:0% advanced applications things like nested variants align:start position:0% nested variants align:start position:0% nested variants some of the spatial yield modeling align:start position:0% some of the spatial yield modeling align:start position:0% some of the spatial yield modeling that sort of thing align:start position:0% align:start position:0% so i think it works well to have three align:start position:0% so i think it works well to have three align:start position:0% so i think it works well to have three or even four members in the team so the align:start position:0% or even four members in the team so the align:start position:0% or even four members in the team so the goal is align:start position:0% goal is align:start position:0% goal is have a fairly rich data set and have align:start position:0% have a fairly rich data set and have align:start position:0% have a fairly rich data set and have multiple members of the team applying align:start position:0% multiple members of the team applying align:start position:0% multiple members of the team applying different kinds of analyses to the align:start position:0% different kinds of analyses to the align:start position:0% different kinds of analyses to the problem align:start position:0% problem align:start position:0% problem okay so align:start position:0% okay so align:start position:0% okay so i think it does work where align:start position:0% i think it does work where align:start position:0% i think it does work where you're able to identify contributions align:start position:0% you're able to identify contributions align:start position:0% you're able to identify contributions from each of the align:start position:0% from each of the align:start position:0% from each of the each of the team members but the whole align:start position:0% each of the team members but the whole align:start position:0% each of the team members but the whole point of this is is actually align:start position:0% point of this is is actually align:start position:0% point of this is is actually you know to be able to align:start position:0% you know to be able to align:start position:0% you know to be able to pick over the data align:start position:0% pick over the data align:start position:0% pick over the data talk about the data align:start position:0% talk about the data align:start position:0% talk about the data talk about your approach to align:start position:0% talk about your approach to align:start position:0% talk about your approach to analysis align:start position:0% analysis align:start position:0% analysis and looking at the problem with your align:start position:0% and looking at the problem with your align:start position:0% and looking at the problem with your teammates align:start position:0% teammates align:start position:0% teammates thursday the project proposal is due align:start position:0% thursday the project proposal is due align:start position:0% thursday the project proposal is due this is just one or two pages describing align:start position:0% this is just one or two pages describing align:start position:0% this is just one or two pages describing the basic data the basic align:start position:0% the basic data the basic align:start position:0% the basic data the basic problem in your plan for attack what align:start position:0% problem in your plan for attack what align:start position:0% problem in your plan for attack what kinds of analyses you hope to do and of align:start position:0% kinds of analyses you hope to do and of align:start position:0% kinds of analyses you hope to do and of course list who your team members are align:start position:0% course list who your team members are align:start position:0% course list who your team members are and what their roles are align:start position:0% and what their roles are align:start position:0% and what their roles are the goal then is over the coming week align:start position:0% the goal then is over the coming week align:start position:0% the goal then is over the coming week and maybe even starting align:start position:0% and maybe even starting align:start position:0% and maybe even starting perhaps even after class for for folks align:start position:0% perhaps even after class for for folks align:start position:0% perhaps even after class for for folks in singapore align:start position:0% in singapore align:start position:0% in singapore on thursday certainly on tuesday align:start position:0% on thursday certainly on tuesday align:start position:0% on thursday certainly on tuesday to have a align:start position:0% to have a align:start position:0% to have a a chance to meet with me align:start position:0% a chance to meet with me align:start position:0% a chance to meet with me or dave if he's available align:start position:0% or dave if he's available align:start position:0% or dave if he's available to basically go over your plan align:start position:0% to basically go over your plan align:start position:0% to basically go over your plan and your analysis so far if you're if align:start position:0% and your analysis so far if you're if align:start position:0% and your analysis so far if you're if you're part way through it so we'll need align:start position:0% you're part way through it so we'll need align:start position:0% you're part way through it so we'll need to do that by video conference with uh align:start position:0% to do that by video conference with uh align:start position:0% to do that by video conference with uh with folks in singapore if we do it align:start position:0% with folks in singapore if we do it align:start position:0% with folks in singapore if we do it right after class we can probably align:start position:0% right after class we can probably align:start position:0% right after class we can probably try to arrange align:start position:0% try to arrange align:start position:0% try to arrange this facility otherwise we'll use the align:start position:0% this facility otherwise we'll use the align:start position:0% this facility otherwise we'll use the conference room align:start position:0% conference room align:start position:0% conference room near the align:start position:0% near the align:start position:0% near the the align:start position:0% the align:start position:0% the sma align:start position:0% sma align:start position:0% sma your sma offices align:start position:0% your sma offices align:start position:0% your sma offices then the expectations are two parts align:start position:0% then the expectations are two parts align:start position:0% then the expectations are two parts there'll be a group presentation align:start position:0% there'll be a group presentation align:start position:0% there'll be a group presentation by the way not necessarily every single align:start position:0% by the way not necessarily every single align:start position:0% by the way not necessarily every single member of your team needs to align:start position:0% member of your team needs to align:start position:0% member of your team needs to present part of the presentation trying align:start position:0% present part of the presentation trying align:start position:0% present part of the presentation trying to have four team members in a 15-minute align:start position:0% to have four team members in a 15-minute align:start position:0% to have four team members in a 15-minute chunk align:start position:0% chunk align:start position:0% chunk often can be kind of a kind of a align:start position:0% often can be kind of a kind of a align:start position:0% often can be kind of a kind of a coordination mess so as a team you need align:start position:0% coordination mess so as a team you need align:start position:0% coordination mess so as a team you need to put together a presentation align:start position:0% to put together a presentation align:start position:0% to put together a presentation and have one or maybe two people uh most align:start position:0% and have one or maybe two people uh most align:start position:0% and have one or maybe two people uh most likely presenting uh some of those align:start position:0% likely presenting uh some of those align:start position:0% likely presenting uh some of those results and then at the end of that week align:start position:0% results and then at the end of that week align:start position:0% results and then at the end of that week on friday uh may 16th which is the align:start position:0% on friday uh may 16th which is the align:start position:0% on friday uh may 16th which is the the end of the term given that we do not align:start position:0% the end of the term given that we do not align:start position:0% the end of the term given that we do not have a final exam align:start position:0% have a final exam align:start position:0% have a final exam a single shared group report align:start position:0% a single shared group report align:start position:0% a single shared group report needs to be submitted align:start position:0% align:start position:0% these are just some example example align:start position:0% these are just some example example align:start position:0% these are just some example example topics i'm not sure if those have been align:start position:0% topics i'm not sure if those have been align:start position:0% topics i'm not sure if those have been drawn from align:start position:0% drawn from align:start position:0% drawn from past semesters or not align:start position:0% past semesters or not align:start position:0% past semesters or not uh here's some additional kinds of align:start position:0% uh here's some additional kinds of align:start position:0% uh here's some additional kinds of suggestions that align:start position:0% suggestions that align:start position:0% suggestions that that that i mentioned kinds of analyses align:start position:0% that that i mentioned kinds of analyses align:start position:0% that that i mentioned kinds of analyses that might be of interest align:start position:0% that might be of interest align:start position:0% that might be of interest uh in the report these are some align:start position:0% uh in the report these are some align:start position:0% uh in the report these are some guidelines on what we'd like to see in align:start position:0% guidelines on what we'd like to see in align:start position:0% guidelines on what we'd like to see in the report one note here and i haven't align:start position:0% the report one note here and i haven't align:start position:0% the report one note here and i haven't posted it yet align:start position:0% posted it yet align:start position:0% posted it yet is whoops is we'd like to align:start position:0% is whoops is we'd like to align:start position:0% is whoops is we'd like to use a common align:start position:0% use a common align:start position:0% use a common template just for the formatting and align:start position:0% template just for the formatting and align:start position:0% template just for the formatting and guidelines and basically the ieee align:start position:0% guidelines and basically the ieee align:start position:0% guidelines and basically the ieee two-column format has worked really align:start position:0% two-column format has worked really align:start position:0% two-column format has worked really nicely we have word templates i think we align:start position:0% nicely we have word templates i think we align:start position:0% nicely we have word templates i think we even have latex templates if there's align:start position:0% even have latex templates if there's align:start position:0% even have latex templates if there's anybody who still uses latex align:start position:0% anybody who still uses latex align:start position:0% anybody who still uses latex i will post that align:start position:0% i will post that align:start position:0% i will post that that template align:start position:0% that template align:start position:0% that template on the web what's nice about that align:start position:0% on the web what's nice about that align:start position:0% on the web what's nice about that is we can actually then put together all align:start position:0% is we can actually then put together all align:start position:0% is we can actually then put together all of the written reports from the class align:start position:0% of the written reports from the class align:start position:0% of the written reports from the class into almost a little booklet and it's a align:start position:0% into almost a little booklet and it's a align:start position:0% into almost a little booklet and it's a it almost looks like a little a little align:start position:0% it almost looks like a little a little align:start position:0% it almost looks like a little a little mini align:start position:0% mini align:start position:0% mini uh journal or a little mini conference align:start position:0% uh journal or a little mini conference align:start position:0% uh journal or a little mini conference proceeding if you will align:start position:0% proceeding if you will align:start position:0% proceeding if you will with these align:start position:0% with these align:start position:0% with these these reports it also is a nice align:start position:0% these reports it also is a nice align:start position:0% these reports it also is a nice professional looking align:start position:0% professional looking align:start position:0% professional looking report so we'll distribute align:start position:0% report so we'll distribute align:start position:0% report so we'll distribute that as well align:start position:0% that as well align:start position:0% that as well so finally i just uh align:start position:0% so finally i just uh align:start position:0% so finally i just uh grabbed from the past two or three years align:start position:0% grabbed from the past two or three years align:start position:0% grabbed from the past two or three years some of the titles from some of the align:start position:0% some of the titles from some of the align:start position:0% some of the titles from some of the process align:start position:0% process align:start position:0% process uh or team report uh align:start position:0% uh or team report uh align:start position:0% uh or team report uh and presentations align:start position:0% and presentations align:start position:0% and presentations just to give you a little bit of a sense align:start position:0% just to give you a little bit of a sense align:start position:0% just to give you a little bit of a sense of the variety of of kinds of things align:start position:0% of the variety of of kinds of things align:start position:0% of the variety of of kinds of things that we saw align:start position:0% that we saw align:start position:0% that we saw so those might give you some idea by the align:start position:0% so those might give you some idea by the align:start position:0% so those might give you some idea by the way back in 2005 we we were more focused align:start position:0% way back in 2005 we we were more focused align:start position:0% way back in 2005 we we were more focused on like two-person projects or align:start position:0% on like two-person projects or align:start position:0% on like two-person projects or one-person projects align:start position:0% one-person projects align:start position:0% one-person projects uh and these were great reports but we align:start position:0% uh and these were great reports but we align:start position:0% uh and these were great reports but we did decide that uh align:start position:0% did decide that uh align:start position:0% did decide that uh sharing and interacting uh made made align:start position:0% sharing and interacting uh made made align:start position:0% sharing and interacting uh made made things uh align:start position:0% things uh align:start position:0% things uh a better learning experience for all so align:start position:0% a better learning experience for all so align:start position:0% a better learning experience for all so so with that let me open it up anybody align:start position:0% so with that let me open it up anybody align:start position:0% so with that let me open it up anybody have questions on how the team reports align:start position:0% have questions on how the team reports align:start position:0% have questions on how the team reports team projects team presentations need to align:start position:0% team projects team presentations need to align:start position:0% team projects team presentations need to go align:start position:0% go align:start position:0% go i'll try to break align:start position:0% i'll try to break align:start position:0% i'll try to break does pretty much everybody have a team align:start position:0% does pretty much everybody have a team align:start position:0% does pretty much everybody have a team in formation align:start position:0% in formation align:start position:0% in formation are there people who kind of are still align:start position:0% are there people who kind of are still align:start position:0% are there people who kind of are still looking for possible partners align:start position:0% align:start position:0% okay one or two okay i'll try to break align:start position:0% okay one or two okay i'll try to break align:start position:0% okay one or two okay i'll try to break you know maybe five minutes earlier align:start position:0% you know maybe five minutes earlier align:start position:0% you know maybe five minutes earlier which may mean actually break on time um align:start position:0% which may mean actually break on time um align:start position:0% which may mean actually break on time um uh align:start position:0% uh align:start position:0% uh at the end of class so align:start position:0% at the end of class so align:start position:0% at the end of class so uh could be a little bit of a discussion align:start position:0% uh could be a little bit of a discussion align:start position:0% uh could be a little bit of a discussion while people are still face to face uh align:start position:0% while people are still face to face uh align:start position:0% while people are still face to face uh on that align:start position:0% on that align:start position:0% on that any other questions on that align:start position:0% align:start position:0% professor yes align:start position:0% professor yes align:start position:0% professor yes um just now you mentioned um align:start position:0% um just now you mentioned um align:start position:0% um just now you mentioned um having a rich set of data align:start position:0% having a rich set of data align:start position:0% having a rich set of data so align:start position:0% so align:start position:0% so from my understanding we are supposed to align:start position:0% from my understanding we are supposed to align:start position:0% from my understanding we are supposed to do the experiments ourselves so i would align:start position:0% do the experiments ourselves so i would align:start position:0% do the experiments ourselves so i would like to know how align:start position:0% like to know how align:start position:0% like to know how big a set of data are you expecting align:start position:0% align:start position:0% well align:start position:0% well align:start position:0% well i align:start position:0% i align:start position:0% i my expectation or align:start position:0% my expectation or align:start position:0% my expectation or minimal expectation i would say is more align:start position:0% minimal expectation i would say is more align:start position:0% minimal expectation i would say is more that you might be able to tap into data align:start position:0% that you might be able to tap into data align:start position:0% that you might be able to tap into data that has already previously been align:start position:0% that has already previously been align:start position:0% that has already previously been generated align:start position:0% generated align:start position:0% generated it's ambitious and i would applaud align:start position:0% it's ambitious and i would applaud align:start position:0% it's ambitious and i would applaud uh people wanting to go and generate align:start position:0% uh people wanting to go and generate align:start position:0% uh people wanting to go and generate additional data align:start position:0% additional data align:start position:0% additional data so we could certainly talk and depending align:start position:0% so we could certainly talk and depending align:start position:0% so we could certainly talk and depending on the data you have in mind but but align:start position:0% on the data you have in mind but but align:start position:0% on the data you have in mind but but what i mean by a rich set of data is align:start position:0% what i mean by a rich set of data is align:start position:0% what i mean by a rich set of data is something that will support align:start position:0% something that will support align:start position:0% something that will support enough of the kinds of analyses that align:start position:0% enough of the kinds of analyses that align:start position:0% enough of the kinds of analyses that we've been been talking about through align:start position:0% we've been been talking about through align:start position:0% we've been been talking about through through the term align:start position:0% through the term align:start position:0% through the term so if it's a doe it needs to be align:start position:0% so if it's a doe it needs to be align:start position:0% so if it's a doe it needs to be you know probably several parameters of align:start position:0% you know probably several parameters of align:start position:0% you know probably several parameters of input several parameters of output align:start position:0% input several parameters of output align:start position:0% input several parameters of output maybe you have some align:start position:0% maybe you have some align:start position:0% maybe you have some additional twist to it that's kind of align:start position:0% additional twist to it that's kind of align:start position:0% additional twist to it that's kind of interesting or or non-standard align:start position:0% interesting or or non-standard align:start position:0% interesting or or non-standard that included some align:start position:0% that included some align:start position:0% that included some some thorny issues in the data maybe align:start position:0% some thorny issues in the data maybe align:start position:0% some thorny issues in the data maybe it's an unbalanced design or an align:start position:0% it's an unbalanced design or an align:start position:0% it's an unbalanced design or an imperfect design and you have to try to align:start position:0% imperfect design and you have to try to align:start position:0% imperfect design and you have to try to try to accommodate that i'm i'm all i align:start position:0% try to accommodate that i'm i'm all i align:start position:0% try to accommodate that i'm i'm all i mean is is uh align:start position:0% align:start position:0% you know align:start position:0% you know align:start position:0% you know enough data to do what you would like to align:start position:0% enough data to do what you would like to align:start position:0% enough data to do what you would like to propose align:start position:0% propose align:start position:0% propose to do and that that that we can talk align:start position:0% to do and that that that we can talk align:start position:0% to do and that that that we can talk about uh align:start position:0% about uh align:start position:0% about uh about finding acceptable align:start position:0% about finding acceptable align:start position:0% about finding acceptable so were you actually looking at going in align:start position:0% so were you actually looking at going in align:start position:0% so were you actually looking at going in and generating some data on a one of the align:start position:0% and generating some data on a one of the align:start position:0% and generating some data on a one of the processes or tools or equipment there align:start position:0% align:start position:0% uh align:start position:0% uh align:start position:0% uh uh my girl haven't had a discussion yet align:start position:0% uh my girl haven't had a discussion yet align:start position:0% uh my girl haven't had a discussion yet um align:start position:0% um align:start position:0% um it's a requirement oh okay yeah so we align:start position:0% it's a requirement oh okay yeah so we align:start position:0% it's a requirement oh okay yeah so we can follow up with hayden align:start position:0% can follow up with hayden align:start position:0% can follow up with hayden on that i think you know bonding and align:start position:0% on that i think you know bonding and align:start position:0% on that i think you know bonding and that sort of thing would be great align:start position:0% align:start position:0% so we won't be too ambitious it'll be align:start position:0% so we won't be too ambitious it'll be align:start position:0% so we won't be too ambitious it'll be doable align:start position:0% align:start position:0% okay other questions align:start position:0% align:start position:0% all kind of manufacturing processes align:start position:0% all kind of manufacturing processes align:start position:0% all kind of manufacturing processes but i have a align:start position:0% but i have a align:start position:0% but i have a past lab data align:start position:0% past lab data align:start position:0% past lab data from kind of material engineering align:start position:0% align:start position:0% sure so the question was you know is is align:start position:0% sure so the question was you know is is align:start position:0% sure so the question was you know is is material engineering data align:start position:0% material engineering data align:start position:0% material engineering data mechanical pro processes mechanical align:start position:0% mechanical pro processes mechanical align:start position:0% mechanical pro processes mechanical properties align:start position:0% properties align:start position:0% properties is that okay and the answer is align:start position:0% is that okay and the answer is align:start position:0% is that okay and the answer is absolutely i mean if you look at some of align:start position:0% absolutely i mean if you look at some of align:start position:0% absolutely i mean if you look at some of these past align:start position:0% these past align:start position:0% these past projects you've got everything from you align:start position:0% projects you've got everything from you align:start position:0% projects you've got everything from you know some things that look kind of like align:start position:0% know some things that look kind of like align:start position:0% know some things that look kind of like semiconductor stuff laser diode it's not align:start position:0% semiconductor stuff laser diode it's not align:start position:0% semiconductor stuff laser diode it's not transistors but you have other things of align:start position:0% transistors but you have other things of align:start position:0% transistors but you have other things of of bending printed circuit boards sheet align:start position:0% of bending printed circuit boards sheet align:start position:0% of bending printed circuit boards sheet metal blanket surface quality and align:start position:0% metal blanket surface quality and align:start position:0% metal blanket surface quality and milling align:start position:0% milling align:start position:0% milling all kinds this is sort of general align:start position:0% all kinds this is sort of general align:start position:0% all kinds this is sort of general general manufacturing align:start position:0% general manufacturing align:start position:0% general manufacturing and general processes align:start position:0% and general processes align:start position:0% and general processes so that would be no problem at all yeah align:start position:0% align:start position:0% yeah align:start position:0% yeah align:start position:0% yeah can you post like one of the reports on align:start position:0% can you post like one of the reports on align:start position:0% can you post like one of the reports on stanley so you can take a look align:start position:0% stanley so you can take a look align:start position:0% stanley so you can take a look uh yeah well i think what we've done in align:start position:0% uh yeah well i think what we've done in align:start position:0% uh yeah well i think what we've done in the past is put up some example some of align:start position:0% the past is put up some example some of align:start position:0% the past is put up some example some of the prior year reports we haven't posted align:start position:0% the prior year reports we haven't posted align:start position:0% the prior year reports we haven't posted those yet but i'll try to do that today align:start position:0% those yet but i'll try to do that today align:start position:0% those yet but i'll try to do that today great thank you align:start position:0% align:start position:0% okay any other questions align:start position:0% okay any other questions align:start position:0% okay any other questions all right let me switch here align:start position:0% align:start position:0% so align:start position:0% so align:start position:0% so again what we're doing now is sort of align:start position:0% again what we're doing now is sort of align:start position:0% again what we're doing now is sort of switching to some case studies there's align:start position:0% switching to some case studies there's align:start position:0% switching to some case studies there's still align:start position:0% still align:start position:0% still perhaps uh in a few of the lectures some align:start position:0% perhaps uh in a few of the lectures some align:start position:0% perhaps uh in a few of the lectures some new material if you will coming along align:start position:0% new material if you will coming along align:start position:0% new material if you will coming along in in this case study the first case align:start position:0% in in this case study the first case align:start position:0% in in this case study the first case study align:start position:0% study align:start position:0% study uh i don't think there's actually that align:start position:0% uh i don't think there's actually that align:start position:0% uh i don't think there's actually that much new in terms of uh align:start position:0% much new in terms of uh align:start position:0% much new in terms of uh methodologies and so on i just think align:start position:0% methodologies and so on i just think align:start position:0% methodologies and so on i just think this is a nice example though a little align:start position:0% this is a nice example though a little align:start position:0% this is a nice example though a little dated now of a fairly align:start position:0% dated now of a fairly align:start position:0% dated now of a fairly straightforward but thorough application align:start position:0% straightforward but thorough application align:start position:0% straightforward but thorough application of design of experiments and response align:start position:0% of design of experiments and response align:start position:0% of design of experiments and response surface modeling align:start position:0% surface modeling align:start position:0% surface modeling so i also by the way posted align:start position:0% so i also by the way posted align:start position:0% so i also by the way posted the paper align:start position:0% the paper align:start position:0% the paper for align:start position:0% for align:start position:0% for this case study on the site as well it's align:start position:0% this case study on the site as well it's align:start position:0% this case study on the site as well it's down in the reading section that goes align:start position:0% down in the reading section that goes align:start position:0% down in the reading section that goes readings that goes with the lecture align:start position:0% readings that goes with the lecture align:start position:0% readings that goes with the lecture so you know you'll have to go and pop align:start position:0% so you know you'll have to go and pop align:start position:0% so you know you'll have to go and pop off this align:start position:0% off this align:start position:0% off this what is it eight or nine page align:start position:0% what is it eight or nine page align:start position:0% what is it eight or nine page paper align:start position:0% paper align:start position:0% paper as well i should mention some of the align:start position:0% as well i should mention some of the align:start position:0% as well i should mention some of the other case studies will also have papers align:start position:0% other case studies will also have papers align:start position:0% other case studies will also have papers from the literature align:start position:0% from the literature align:start position:0% from the literature for you to grab align:start position:0% for you to grab align:start position:0% for you to grab and read align:start position:0% and read align:start position:0% and read so this case study comes from align:start position:0% so this case study comes from align:start position:0% so this case study comes from semiconductor manufacturing align:start position:0% semiconductor manufacturing align:start position:0% semiconductor manufacturing it's a little dated now it goes back to align:start position:0% it's a little dated now it goes back to align:start position:0% it's a little dated now it goes back to align:start position:0% align:start position:0% but this was actually kind of just the align:start position:0% but this was actually kind of just the align:start position:0% but this was actually kind of just the time when things like align:start position:0% time when things like align:start position:0% time when things like statistical process control more align:start position:0% statistical process control more align:start position:0% statistical process control more thorough design of experiments response align:start position:0% thorough design of experiments response align:start position:0% thorough design of experiments response service modeling was really finally align:start position:0% service modeling was really finally align:start position:0% service modeling was really finally hitting the semiconductor align:start position:0% hitting the semiconductor align:start position:0% hitting the semiconductor industry align:start position:0% industry align:start position:0% industry and really align:start position:0% and really align:start position:0% and really starting to be more align:start position:0% starting to be more align:start position:0% starting to be more widespread in align:start position:0% widespread in align:start position:0% widespread in adoption align:start position:0% adoption align:start position:0% adoption so i think this is a nice case it's a align:start position:0% so i think this is a nice case it's a align:start position:0% so i think this is a nice case it's a nice case study because they actually align:start position:0% nice case study because they actually align:start position:0% nice case study because they actually show us the align:start position:0% show us the align:start position:0% show us the experimental conditions the actual data align:start position:0% experimental conditions the actual data align:start position:0% experimental conditions the actual data and the actual models that they align:start position:0% and the actual models that they align:start position:0% and the actual models that they generated align:start position:0% generated align:start position:0% generated some of the things are a little dated align:start position:0% some of the things are a little dated align:start position:0% some of the things are a little dated for example they talk about align:start position:0% for example they talk about align:start position:0% for example they talk about millimeter wafers align:start position:0% millimeter wafers align:start position:0% millimeter wafers so this is five inch wafer processing align:start position:0% so this is five inch wafer processing align:start position:0% so this is five inch wafer processing which is kind of a weird size but it align:start position:0% which is kind of a weird size but it align:start position:0% which is kind of a weird size but it also you know we're up to 300 millimeter align:start position:0% also you know we're up to 300 millimeter align:start position:0% also you know we're up to 300 millimeter 12 inch wafers so it shows the march of align:start position:0% 12 inch wafers so it shows the march of align:start position:0% 12 inch wafers so it shows the march of progress of of technology however align:start position:0% progress of of technology however align:start position:0% progress of of technology however cbd and cvd of blanket tungsten is is align:start position:0% cbd and cvd of blanket tungsten is is align:start position:0% cbd and cvd of blanket tungsten is is still heavily used very integral so it's align:start position:0% still heavily used very integral so it's align:start position:0% still heavily used very integral so it's still a relevant process align:start position:0% still a relevant process align:start position:0% still a relevant process so what i want to do align:start position:0% so what i want to do align:start position:0% so what i want to do in looking at this case is go through align:start position:0% in looking at this case is go through align:start position:0% in looking at this case is go through talk a little bit about the bank align:start position:0% talk a little bit about the bank align:start position:0% talk a little bit about the bank background of tungsten cbd things that align:start position:0% background of tungsten cbd things that align:start position:0% background of tungsten cbd things that aren't in the paper but may give you a align:start position:0% aren't in the paper but may give you a align:start position:0% aren't in the paper but may give you a little bit of a align:start position:0% little bit of a align:start position:0% little bit of a a feel on the more of a feel on the align:start position:0% a feel on the more of a feel on the align:start position:0% a feel on the more of a feel on the process align:start position:0% process align:start position:0% process talk a little bit about the preliminary align:start position:0% talk a little bit about the preliminary align:start position:0% talk a little bit about the preliminary work that align:start position:0% work that align:start position:0% work that they did align:start position:0% they did align:start position:0% they did looking back at the prior paper that align:start position:0% looking back at the prior paper that align:start position:0% looking back at the prior paper that they align:start position:0% they align:start position:0% they that they cite align:start position:0% that they cite align:start position:0% that they cite and then we'll go through the align:start position:0% and then we'll go through the align:start position:0% and then we'll go through the experimental design that they performed align:start position:0% experimental design that they performed align:start position:0% experimental design that they performed to try to optimize align:start position:0% to try to optimize align:start position:0% to try to optimize this tungsten cbd process to understand align:start position:0% this tungsten cbd process to understand align:start position:0% this tungsten cbd process to understand it model it and then optimize it so align:start position:0% it model it and then optimize it so align:start position:0% it model it and then optimize it so we'll look at the data we'll do some rsm align:start position:0% we'll look at the data we'll do some rsm align:start position:0% we'll look at the data we'll do some rsm analysis or align:start position:0% analysis or align:start position:0% analysis or mostly looking at align:start position:0% mostly looking at align:start position:0% mostly looking at their analysis in the paper align:start position:0% their analysis in the paper align:start position:0% their analysis in the paper but align:start position:0% but align:start position:0% but assuming i have time at the end and i align:start position:0% assuming i have time at the end and i align:start position:0% assuming i have time at the end and i think i will align:start position:0% think i will align:start position:0% think i will i'll switch over and we'll we'll use align:start position:0% i'll switch over and we'll we'll use align:start position:0% i'll switch over and we'll we'll use play around with jump a little bit with align:start position:0% play around with jump a little bit with align:start position:0% play around with jump a little bit with that same data set and align:start position:0% that same data set and align:start position:0% that same data set and show you a few of the kinds of analyses align:start position:0% show you a few of the kinds of analyses align:start position:0% show you a few of the kinds of analyses that are possible align:start position:0% that are possible align:start position:0% that are possible with that tool align:start position:0% with that tool align:start position:0% with that tool both for align:start position:0% both for align:start position:0% both for regression some of the stepwise align:start position:0% regression some of the stepwise align:start position:0% regression some of the stepwise regression that's referred to align:start position:0% regression that's referred to align:start position:0% regression that's referred to in the paper and then some of the neat align:start position:0% in the paper and then some of the neat align:start position:0% in the paper and then some of the neat tools actually in the interactive align:start position:0% tools actually in the interactive align:start position:0% tools actually in the interactive capabilities of jump for for align:start position:0% capabilities of jump for for align:start position:0% capabilities of jump for for optimization align:start position:0% optimization align:start position:0% optimization now throughout this align:start position:0% now throughout this align:start position:0% now throughout this think a little bit about where you see align:start position:0% think a little bit about where you see align:start position:0% think a little bit about where you see weaknesses or alternative ideas or align:start position:0% weaknesses or alternative ideas or align:start position:0% weaknesses or alternative ideas or things like that because at the end i align:start position:0% things like that because at the end i align:start position:0% things like that because at the end i might ask what are some of your ideas of align:start position:0% might ask what are some of your ideas of align:start position:0% might ask what are some of your ideas of things you think might have been align:start position:0% things you think might have been align:start position:0% things you think might have been interesting align:start position:0% interesting align:start position:0% interesting uh or important to try or to improve on align:start position:0% uh or important to try or to improve on align:start position:0% uh or important to try or to improve on this so have that in mind what what align:start position:0% this so have that in mind what what align:start position:0% this so have that in mind what what where where do you think they did a good align:start position:0% where where do you think they did a good align:start position:0% where where do you think they did a good job what do you think they did a a align:start position:0% job what do you think they did a a align:start position:0% job what do you think they did a a weaker job what might you suggest and align:start position:0% weaker job what might you suggest and align:start position:0% weaker job what might you suggest and i'll ask you that question at the at the align:start position:0% i'll ask you that question at the at the align:start position:0% i'll ask you that question at the at the end okay align:start position:0% end okay align:start position:0% end okay so let me talk a little bit about align:start position:0% so let me talk a little bit about align:start position:0% so let me talk a little bit about tungsten and in particular tungsten align:start position:0% tungsten and in particular tungsten align:start position:0% tungsten and in particular tungsten plugs align:start position:0% plugs align:start position:0% plugs so align:start position:0% so align:start position:0% so tungsten align:start position:0% tungsten align:start position:0% tungsten here is generally used as a conductor in align:start position:0% here is generally used as a conductor in align:start position:0% here is generally used as a conductor in semiconductor fabrication it's not a align:start position:0% semiconductor fabrication it's not a align:start position:0% semiconductor fabrication it's not a great conductor meaning it it's align:start position:0% great conductor meaning it it's align:start position:0% great conductor meaning it it's often a little bit higher resistivity align:start position:0% often a little bit higher resistivity align:start position:0% often a little bit higher resistivity than other metals like aluminum align:start position:0% than other metals like aluminum align:start position:0% than other metals like aluminum or align:start position:0% or align:start position:0% or or now copper align:start position:0% or now copper align:start position:0% or now copper it's align:start position:0% it's align:start position:0% it's continues to be heavily used for align:start position:0% continues to be heavily used for align:start position:0% continues to be heavily used for contacts and plugs align:start position:0% contacts and plugs align:start position:0% contacts and plugs so contacts are the holes that are align:start position:0% so contacts are the holes that are align:start position:0% so contacts are the holes that are drilled between or into align:start position:0% drilled between or into align:start position:0% drilled between or into an insulating layer from the first level align:start position:0% an insulating layer from the first level align:start position:0% an insulating layer from the first level of metal down to the underlying silicon align:start position:0% of metal down to the underlying silicon align:start position:0% of metal down to the underlying silicon okay align:start position:0% okay align:start position:0% okay plugs are the same idea but at other align:start position:0% plugs are the same idea but at other align:start position:0% plugs are the same idea but at other levels of interconnect so you would have align:start position:0% levels of interconnect so you would have align:start position:0% levels of interconnect so you would have a a plug align:start position:0% a a plug align:start position:0% a a plug from metal one to metal two or metal two align:start position:0% from metal one to metal two or metal two align:start position:0% from metal one to metal two or metal two to metal 3 and so on align:start position:0% to metal 3 and so on align:start position:0% to metal 3 and so on terminology is not completely uniform align:start position:0% terminology is not completely uniform align:start position:0% terminology is not completely uniform but that's that's align:start position:0% but that's that's align:start position:0% but that's that's that's the typical terminology align:start position:0% align:start position:0% the examples drawn from from this paper align:start position:0% the examples drawn from from this paper align:start position:0% the examples drawn from from this paper they talk about tungsten plugs in align:start position:0% they talk about tungsten plugs in align:start position:0% they talk about tungsten plugs in conjunction with aluminum interconnect align:start position:0% conjunction with aluminum interconnect align:start position:0% conjunction with aluminum interconnect and they're actually mostly focused i align:start position:0% and they're actually mostly focused i align:start position:0% and they're actually mostly focused i think on tungsten between align:start position:0% think on tungsten between align:start position:0% think on tungsten between or tungsten contacts align:start position:0% or tungsten contacts align:start position:0% or tungsten contacts with the idea that in older technology align:start position:0% with the idea that in older technology align:start position:0% with the idea that in older technology in fact you didn't need to drill and align:start position:0% in fact you didn't need to drill and align:start position:0% in fact you didn't need to drill and form separate align:start position:0% form separate align:start position:0% form separate plugs between layers you could just align:start position:0% plugs between layers you could just align:start position:0% plugs between layers you could just allow say an aluminum deposition to go align:start position:0% allow say an aluminum deposition to go align:start position:0% allow say an aluminum deposition to go in and fill align:start position:0% in and fill align:start position:0% in and fill that layer and then align:start position:0% that layer and then align:start position:0% that layer and then etch off where you where you didn't want align:start position:0% etch off where you where you didn't want align:start position:0% etch off where you where you didn't want it align:start position:0% it align:start position:0% it with scaling it became pretty much align:start position:0% with scaling it became pretty much align:start position:0% with scaling it became pretty much essential to be able to form the plugs align:start position:0% essential to be able to form the plugs align:start position:0% essential to be able to form the plugs separately from the interconnect align:start position:0% separately from the interconnect align:start position:0% separately from the interconnect so align:start position:0% so align:start position:0% so then tungsten was used to fill fill align:start position:0% then tungsten was used to fill fill align:start position:0% then tungsten was used to fill fill those plugs between metal layers as well align:start position:0% those plugs between metal layers as well align:start position:0% those plugs between metal layers as well as we've gone to copper interconnect align:start position:0% as we've gone to copper interconnect align:start position:0% as we've gone to copper interconnect tungsten still appears align:start position:0% tungsten still appears align:start position:0% tungsten still appears in most uh advanced technologies now align:start position:0% in most uh advanced technologies now align:start position:0% in most uh advanced technologies now primarily just as the first level of of align:start position:0% primarily just as the first level of of align:start position:0% primarily just as the first level of of contact align:start position:0% contact align:start position:0% contact from a copper interconnect down to the align:start position:0% from a copper interconnect down to the align:start position:0% from a copper interconnect down to the silicon so it is still used align:start position:0% silicon so it is still used align:start position:0% silicon so it is still used and extensions where it may not purely align:start position:0% and extensions where it may not purely align:start position:0% and extensions where it may not purely be be tungsten plugs but it may also be align:start position:0% be be tungsten plugs but it may also be align:start position:0% be be tungsten plugs but it may also be in some cases align:start position:0% in some cases align:start position:0% in some cases some local align:start position:0% some local align:start position:0% some local some local interconnect align:start position:0% some local interconnect align:start position:0% some local interconnect so you can down very close to the align:start position:0% so you can down very close to the align:start position:0% so you can down very close to the silicon form little straps or other align:start position:0% silicon form little straps or other align:start position:0% silicon form little straps or other sorts of local tungsten based uh align:start position:0% sorts of local tungsten based uh align:start position:0% sorts of local tungsten based uh short very short run align:start position:0% short very short run align:start position:0% short very short run interconnect align:start position:0% interconnect align:start position:0% interconnect to form electrical connections align:start position:0% to form electrical connections align:start position:0% to form electrical connections now the important part of tungsten why align:start position:0% now the important part of tungsten why align:start position:0% now the important part of tungsten why it's used is that it's also align:start position:0% it's used is that it's also align:start position:0% it's used is that it's also can be align:start position:0% can be align:start position:0% can be conformally align:start position:0% conformally align:start position:0% conformally filled align:start position:0% filled align:start position:0% filled and can align:start position:0% and can align:start position:0% and can especially with uh chemical vapor align:start position:0% especially with uh chemical vapor align:start position:0% especially with uh chemical vapor deposition or plasma processing uh can align:start position:0% deposition or plasma processing uh can align:start position:0% deposition or plasma processing uh can fill fairly small holes align:start position:0% fill fairly small holes align:start position:0% fill fairly small holes so the basic idea align:start position:0% so the basic idea align:start position:0% so the basic idea or progression align:start position:0% or progression align:start position:0% or progression in a tungsten deposition align:start position:0% in a tungsten deposition align:start position:0% in a tungsten deposition is you primarily get align:start position:0% is you primarily get align:start position:0% is you primarily get conformal deposition so in time align:start position:0% conformal deposition so in time align:start position:0% conformal deposition so in time the align:start position:0% the align:start position:0% the gases align:start position:0% gases align:start position:0% gases that are that are present in the plasma align:start position:0% that are that are present in the plasma align:start position:0% that are that are present in the plasma chamber and reactor align:start position:0% chamber and reactor align:start position:0% chamber and reactor interface with the surface form align:start position:0% interface with the surface form align:start position:0% interface with the surface form reactions align:start position:0% reactions align:start position:0% reactions and basically build up in a conformal align:start position:0% and basically build up in a conformal align:start position:0% and basically build up in a conformal way to fill align:start position:0% align:start position:0% fill the plug align:start position:0% fill the plug align:start position:0% fill the plug now you want to do this in a way that align:start position:0% now you want to do this in a way that align:start position:0% now you want to do this in a way that does not result in seams or align:start position:0% does not result in seams or align:start position:0% does not result in seams or or voids in the middle so that can be a align:start position:0% or voids in the middle so that can be a align:start position:0% or voids in the middle so that can be a little bit tricky but but that can be align:start position:0% little bit tricky but but that can be align:start position:0% little bit tricky but but that can be done in in such a way that one align:start position:0% done in in such a way that one align:start position:0% done in in such a way that one avoids those those plugs align:start position:0% align:start position:0% now the way this paper uses tungsten align:start position:0% now the way this paper uses tungsten align:start position:0% now the way this paper uses tungsten plugs is with a process they refer to as align:start position:0% plugs is with a process they refer to as align:start position:0% plugs is with a process they refer to as plasma etchback so if i do that align:start position:0% plasma etchback so if i do that align:start position:0% plasma etchback so if i do that conformal deposition align:start position:0% conformal deposition align:start position:0% conformal deposition finally to this this surface align:start position:0% finally to this this surface align:start position:0% finally to this this surface uh now i've got align:start position:0% uh now i've got align:start position:0% uh now i've got you know the align:start position:0% you know the align:start position:0% you know the tungsten not only in the plug but i've align:start position:0% tungsten not only in the plug but i've align:start position:0% tungsten not only in the plug but i've also got all this excess tungsten out align:start position:0% also got all this excess tungsten out align:start position:0% also got all this excess tungsten out here in the align:start position:0% here in the align:start position:0% here in the neighboring regions that i want to align:start position:0% neighboring regions that i want to align:start position:0% neighboring regions that i want to remove align:start position:0% remove align:start position:0% remove and the idea of of etchback was that align:start position:0% and the idea of of etchback was that align:start position:0% and the idea of of etchback was that through the deposition process i fill align:start position:0% through the deposition process i fill align:start position:0% through the deposition process i fill the surface align:start position:0% the surface align:start position:0% the surface but i leave essentially a nearly planar align:start position:0% but i leave essentially a nearly planar align:start position:0% but i leave essentially a nearly planar surface that now if i just etch align:start position:0% surface that now if i just etch align:start position:0% surface that now if i just etch uniformly align:start position:0% uniformly align:start position:0% uniformly back that whole surface align:start position:0% back that whole surface align:start position:0% back that whole surface i can remove down to the underlying align:start position:0% i can remove down to the underlying align:start position:0% i can remove down to the underlying interface metal or interface layer or align:start position:0% interface metal or interface layer or align:start position:0% interface metal or interface layer or the oxide align:start position:0% the oxide align:start position:0% the oxide and leave the metal just in the in the align:start position:0% and leave the metal just in the in the align:start position:0% and leave the metal just in the in the plug where i wanted it align:start position:0% plug where i wanted it align:start position:0% plug where i wanted it in the paper they refer to using this as align:start position:0% in the paper they refer to using this as align:start position:0% in the paper they refer to using this as are using plasma etchback to do this align:start position:0% are using plasma etchback to do this align:start position:0% are using plasma etchback to do this uniform etchback align:start position:0% uniform etchback align:start position:0% uniform etchback in modern processing this step has align:start position:0% in modern processing this step has align:start position:0% in modern processing this step has pretty much been replaced with cmp align:start position:0% pretty much been replaced with cmp align:start position:0% pretty much been replaced with cmp because that has a much higher align:start position:0% because that has a much higher align:start position:0% because that has a much higher controllability and align:start position:0% controllability and align:start position:0% controllability and gets around an important limitation align:start position:0% gets around an important limitation align:start position:0% gets around an important limitation that's kind of pictured here in this in align:start position:0% that's kind of pictured here in this in align:start position:0% that's kind of pictured here in this in this in this diagram align:start position:0% this in this diagram align:start position:0% this in this diagram having to do with align:start position:0% having to do with align:start position:0% having to do with very tight constraints on uniformity align:start position:0% very tight constraints on uniformity align:start position:0% very tight constraints on uniformity with a plasma etchback align:start position:0% with a plasma etchback align:start position:0% with a plasma etchback the basic issue with uniformity is if i align:start position:0% the basic issue with uniformity is if i align:start position:0% the basic issue with uniformity is if i in fact have deposition align:start position:0% in fact have deposition align:start position:0% in fact have deposition that results in different thicknesses align:start position:0% that results in different thicknesses align:start position:0% that results in different thicknesses either in different locations within the align:start position:0% either in different locations within the align:start position:0% either in different locations within the chip align:start position:0% chip align:start position:0% chip or in different chips across the wafer align:start position:0% or in different chips across the wafer align:start position:0% or in different chips across the wafer and i uniformly etch back align:start position:0% and i uniformly etch back align:start position:0% and i uniformly etch back first off if i only etch back this align:start position:0% first off if i only etch back this align:start position:0% first off if i only etch back this amount align:start position:0% amount align:start position:0% amount in both regions then i have not align:start position:0% in both regions then i have not align:start position:0% in both regions then i have not completely removed what i needed to in align:start position:0% completely removed what i needed to in align:start position:0% completely removed what i needed to in the other regions so i have to make sure align:start position:0% the other regions so i have to make sure align:start position:0% the other regions so i have to make sure i do enough overarch align:start position:0% i do enough overarch align:start position:0% i do enough overarch to at least clear off and make sure i align:start position:0% to at least clear off and make sure i align:start position:0% to at least clear off and make sure i don't have any shorts align:start position:0% don't have any shorts align:start position:0% don't have any shorts anywhere on the wafer and when you do align:start position:0% anywhere on the wafer and when you do align:start position:0% anywhere on the wafer and when you do that to make sure that i'm down to the align:start position:0% that to make sure that i'm down to the align:start position:0% that to make sure that i'm down to the surface align:start position:0% surface align:start position:0% surface in the thickest area you get over etch align:start position:0% in the thickest area you get over etch align:start position:0% in the thickest area you get over etch lossage align:start position:0% lossage align:start position:0% lossage in align:start position:0% in align:start position:0% in the thinnest regions so you have a align:start position:0% the thinnest regions so you have a align:start position:0% the thinnest regions so you have a fairly tight constraint then on what the align:start position:0% fairly tight constraint then on what the align:start position:0% fairly tight constraint then on what the thickness uniformity is align:start position:0% thickness uniformity is align:start position:0% thickness uniformity is as well as the etch uniformity in order align:start position:0% as well as the etch uniformity in order align:start position:0% as well as the etch uniformity in order to achieve that align:start position:0% to achieve that align:start position:0% to achieve that okay so the paper does talk a bit and in align:start position:0% okay so the paper does talk a bit and in align:start position:0% okay so the paper does talk a bit and in fact we'll see later explicitly models align:start position:0% fact we'll see later explicitly models align:start position:0% fact we'll see later explicitly models the wafer uniformity as one of the align:start position:0% the wafer uniformity as one of the align:start position:0% the wafer uniformity as one of the outputs align:start position:0% outputs align:start position:0% outputs in order to try to understand as they align:start position:0% in order to try to understand as they align:start position:0% in order to try to understand as they play with some of the process conditions align:start position:0% play with some of the process conditions align:start position:0% play with some of the process conditions how uniformity is going to be affected align:start position:0% how uniformity is going to be affected align:start position:0% how uniformity is going to be affected by their choice of these conditions align:start position:0% by their choice of these conditions align:start position:0% by their choice of these conditions okay align:start position:0% okay align:start position:0% okay uh there's a couple of other interesting align:start position:0% uh there's a couple of other interesting align:start position:0% uh there's a couple of other interesting aspects of plasma etchback that are align:start position:0% aspects of plasma etchback that are align:start position:0% aspects of plasma etchback that are that are important uh align:start position:0% that are important uh align:start position:0% that are important uh as well in particular help align:start position:0% align:start position:0% inform the modeling that this paper uh align:start position:0% inform the modeling that this paper uh align:start position:0% inform the modeling that this paper uh undertakes and that has to do with align:start position:0% undertakes and that has to do with align:start position:0% undertakes and that has to do with surface roughness so not only if will align:start position:0% surface roughness so not only if will align:start position:0% surface roughness so not only if will these sorts of features this topography align:start position:0% these sorts of features this topography align:start position:0% these sorts of features this topography like that little dimple there align:start position:0% like that little dimple there align:start position:0% like that little dimple there but also align:start position:0% but also align:start position:0% but also if the surface of the film the deposited align:start position:0% if the surface of the film the deposited align:start position:0% if the surface of the film the deposited film is rough align:start position:0% film is rough align:start position:0% film is rough and i do a uniform plasma etch back align:start position:0% and i do a uniform plasma etch back align:start position:0% and i do a uniform plasma etch back sort of a vertical plasma etch back align:start position:0% sort of a vertical plasma etch back align:start position:0% sort of a vertical plasma etch back that surface roughness can in fact then align:start position:0% that surface roughness can in fact then align:start position:0% that surface roughness can in fact then this is rough align:start position:0% this is rough align:start position:0% this is rough that align:start position:0% that align:start position:0% that in truth may be transmitted to the final align:start position:0% in truth may be transmitted to the final align:start position:0% in truth may be transmitted to the final surface of the of the plug align:start position:0% surface of the of the plug align:start position:0% surface of the of the plug so another important parameter that they align:start position:0% so another important parameter that they align:start position:0% so another important parameter that they will be explicitly modeling is align:start position:0% will be explicitly modeling is align:start position:0% will be explicitly modeling is tungsten the deposited tungsten align:start position:0% tungsten the deposited tungsten align:start position:0% tungsten the deposited tungsten roughness align:start position:0% roughness align:start position:0% roughness okay and generally less rough is better align:start position:0% okay and generally less rough is better align:start position:0% okay and generally less rough is better it's going to give a high better surface align:start position:0% it's going to give a high better surface align:start position:0% it's going to give a high better surface quality less prone to defects more align:start position:0% quality less prone to defects more align:start position:0% quality less prone to defects more uniform in resistance and so on align:start position:0% uniform in resistance and so on align:start position:0% uniform in resistance and so on okay align:start position:0% okay align:start position:0% okay that by the way is another reason for align:start position:0% that by the way is another reason for align:start position:0% that by the way is another reason for switching to cmp align:start position:0% switching to cmp align:start position:0% switching to cmp cmp align:start position:0% cmp align:start position:0% cmp planarizes align:start position:0% planarizes align:start position:0% planarizes can deal with not only this kind of align:start position:0% can deal with not only this kind of align:start position:0% can deal with not only this kind of surface topography align:start position:0% surface topography align:start position:0% surface topography on a large scale align:start position:0% on a large scale align:start position:0% on a large scale but it also smooths the surface on sort align:start position:0% but it also smooths the surface on sort align:start position:0% but it also smooths the surface on sort of the atomic scale and so you tend to align:start position:0% of the atomic scale and so you tend to align:start position:0% of the atomic scale and so you tend to get much align:start position:0% get much align:start position:0% get much much align:start position:0% align:start position:0% higher quality final align:start position:0% higher quality final align:start position:0% higher quality final final plugs align:start position:0% final plugs align:start position:0% final plugs okay any questions on basic tungsten align:start position:0% okay any questions on basic tungsten align:start position:0% okay any questions on basic tungsten little lesson on tungsten plugs so let's align:start position:0% little lesson on tungsten plugs so let's align:start position:0% little lesson on tungsten plugs so let's talk a little bit about the process align:start position:0% talk a little bit about the process align:start position:0% talk a little bit about the process equipment and the process parameters align:start position:0% equipment and the process parameters align:start position:0% equipment and the process parameters that are going to be in in this case align:start position:0% that are going to be in in this case align:start position:0% that are going to be in in this case study align:start position:0% study align:start position:0% study this is a schematic crocs section of an align:start position:0% this is a schematic crocs section of an align:start position:0% this is a schematic crocs section of an applied materials precision 5000 reactor align:start position:0% applied materials precision 5000 reactor align:start position:0% applied materials precision 5000 reactor configured for tungsten cbd align:start position:0% configured for tungsten cbd align:start position:0% configured for tungsten cbd by the way this is a classic tool align:start position:0% by the way this is a classic tool align:start position:0% by the way this is a classic tool the eight uh align:start position:0% the eight uh align:start position:0% the eight uh p5000 align:start position:0% p5000 align:start position:0% p5000 a map p5000 there's one of these in the align:start position:0% a map p5000 there's one of these in the align:start position:0% a map p5000 there's one of these in the smithsonian institute align:start position:0% smithsonian institute align:start position:0% smithsonian institute it was the tool that made applied align:start position:0% it was the tool that made applied align:start position:0% it was the tool that made applied materials what it is today align:start position:0% materials what it is today align:start position:0% materials what it is today it was one of the first single wafer align:start position:0% it was one of the first single wafer align:start position:0% it was one of the first single wafer processing tools that really gave align:start position:0% processing tools that really gave align:start position:0% processing tools that really gave good control over the atmospheric align:start position:0% good control over the atmospheric align:start position:0% good control over the atmospheric conditions for deposition for etching align:start position:0% conditions for deposition for etching align:start position:0% conditions for deposition for etching these sorts of things so so align:start position:0% these sorts of things so so align:start position:0% these sorts of things so so we're seeing a glimpse back into history align:start position:0% we're seeing a glimpse back into history align:start position:0% we're seeing a glimpse back into history here even looking at this cross-section align:start position:0% here even looking at this cross-section align:start position:0% here even looking at this cross-section by the way p5000s are still in use all align:start position:0% by the way p5000s are still in use all align:start position:0% by the way p5000s are still in use all over the place align:start position:0% over the place align:start position:0% over the place we have some align:start position:0% we have some align:start position:0% we have some i guess just one in mtl align:start position:0% i guess just one in mtl align:start position:0% i guess just one in mtl i think mostly for etching align:start position:0% i think mostly for etching align:start position:0% i think mostly for etching but uh even in some fabs i i still run align:start position:0% but uh even in some fabs i i still run align:start position:0% but uh even in some fabs i i still run into them align:start position:0% into them align:start position:0% into them they've been scaled up and improved and align:start position:0% they've been scaled up and improved and align:start position:0% they've been scaled up and improved and configured you know changed in align:start position:0% configured you know changed in align:start position:0% configured you know changed in configuration uh but but they are still align:start position:0% configuration uh but but they are still align:start position:0% configuration uh but but they are still around so the basic idea here is that align:start position:0% around so the basic idea here is that align:start position:0% around so the basic idea here is that the wafer align:start position:0% the wafer align:start position:0% the wafer will sit align:start position:0% will sit align:start position:0% will sit here on a susceptor align:start position:0% here on a susceptor align:start position:0% here on a susceptor in this plasma reactor align:start position:0% in this plasma reactor align:start position:0% in this plasma reactor this low low pressure uh align:start position:0% this low low pressure uh align:start position:0% this low low pressure uh plasma reactor in fact there are usually align:start position:0% plasma reactor in fact there are usually align:start position:0% plasma reactor in fact there are usually pins uh i think three or four pins that align:start position:0% pins uh i think three or four pins that align:start position:0% pins uh i think three or four pins that hold the wafer the robot loads the wafer align:start position:0% hold the wafer the robot loads the wafer align:start position:0% hold the wafer the robot loads the wafer in onto these pins and then the pins uh align:start position:0% in onto these pins and then the pins uh align:start position:0% in onto these pins and then the pins uh these lift fingers here i guess align:start position:0% these lift fingers here i guess align:start position:0% these lift fingers here i guess recede and the wafer then sits on this align:start position:0% recede and the wafer then sits on this align:start position:0% recede and the wafer then sits on this on this susceptor the susceptor align:start position:0% on this susceptor the susceptor align:start position:0% on this susceptor the susceptor is align:start position:0% is align:start position:0% is optically heated so there are these heat align:start position:0% optically heated so there are these heat align:start position:0% optically heated so there are these heat lamps i think there's something like align:start position:0% lamps i think there's something like align:start position:0% lamps i think there's something like thousand watt align:start position:0% thousand watt align:start position:0% thousand watt heat lamps in this configuration that align:start position:0% heat lamps in this configuration that align:start position:0% heat lamps in this configuration that heat up align:start position:0% heat up align:start position:0% heat up and hold the temperature align:start position:0% and hold the temperature align:start position:0% and hold the temperature of the whole platen the whole susceptor align:start position:0% align:start position:0% there's then gas flow coming in from a a align:start position:0% there's then gas flow coming in from a a align:start position:0% there's then gas flow coming in from a a gas feed there's a little mixing chamber align:start position:0% gas feed there's a little mixing chamber align:start position:0% gas feed there's a little mixing chamber in here so that you can mix different align:start position:0% in here so that you can mix different align:start position:0% in here so that you can mix different combinations of different gas flows and align:start position:0% combinations of different gas flows and align:start position:0% combinations of different gas flows and then it goes through a shower head align:start position:0% then it goes through a shower head align:start position:0% then it goes through a shower head this thing right here is typically align:start position:0% this thing right here is typically align:start position:0% this thing right here is typically referred to as a shower head align:start position:0% referred to as a shower head align:start position:0% referred to as a shower head and you can imagine it's basically you align:start position:0% and you can imagine it's basically you align:start position:0% and you can imagine it's basically you know like a shower head in your shower align:start position:0% know like a shower head in your shower align:start position:0% know like a shower head in your shower it's got holes in it that basically align:start position:0% it's got holes in it that basically align:start position:0% it's got holes in it that basically allow the gases to flow through align:start position:0% allow the gases to flow through align:start position:0% allow the gases to flow through hopefully in a fairly uniform align:start position:0% hopefully in a fairly uniform align:start position:0% hopefully in a fairly uniform distribution align:start position:0% distribution align:start position:0% distribution across the surface of the wafer align:start position:0% across the surface of the wafer align:start position:0% across the surface of the wafer and then there's a vacuum control and a align:start position:0% and then there's a vacuum control and a align:start position:0% and then there's a vacuum control and a number of different vacuum outlets align:start position:0% number of different vacuum outlets align:start position:0% number of different vacuum outlets for align:start position:0% for align:start position:0% for for the for the gases to flow align:start position:0% for the for the gases to flow align:start position:0% for the for the gases to flow so some of the key inputs there's lots align:start position:0% so some of the key inputs there's lots align:start position:0% so some of the key inputs there's lots and lots of knobs to control on these align:start position:0% and lots of knobs to control on these align:start position:0% and lots of knobs to control on these equipment align:start position:0% equipment align:start position:0% equipment some of the key inputs that are used in align:start position:0% some of the key inputs that are used in align:start position:0% some of the key inputs that are used in this study align:start position:0% this study align:start position:0% this study in the paper that we're looking at are align:start position:0% in the paper that we're looking at are align:start position:0% in the paper that we're looking at are gap space temperature h2 pressure and align:start position:0% gap space temperature h2 pressure and align:start position:0% gap space temperature h2 pressure and wf6 align:start position:0% wf6 align:start position:0% wf6 pressure so the gap space refers to align:start position:0% pressure so the gap space refers to align:start position:0% pressure so the gap space refers to i believe essentially the height of this align:start position:0% i believe essentially the height of this align:start position:0% i believe essentially the height of this deceptor one can play with the distance align:start position:0% deceptor one can play with the distance align:start position:0% deceptor one can play with the distance between the gas inflow and the wafer align:start position:0% between the gas inflow and the wafer align:start position:0% between the gas inflow and the wafer moving it between say two to four or align:start position:0% moving it between say two to four or align:start position:0% moving it between say two to four or five inches align:start position:0% five inches align:start position:0% five inches and that changes a number of the plasma align:start position:0% and that changes a number of the plasma align:start position:0% and that changes a number of the plasma characteristics it changes perhaps some align:start position:0% characteristics it changes perhaps some align:start position:0% characteristics it changes perhaps some of the gas flow characteristics may also align:start position:0% of the gas flow characteristics may also align:start position:0% of the gas flow characteristics may also change some of the thermal align:start position:0% change some of the thermal align:start position:0% change some of the thermal characteristics so it is a control align:start position:0% characteristics so it is a control align:start position:0% characteristics so it is a control parameter that is not entirely obvious align:start position:0% parameter that is not entirely obvious align:start position:0% parameter that is not entirely obvious up front how it affects the growth of or align:start position:0% up front how it affects the growth of or align:start position:0% up front how it affects the growth of or deposition of this this tungsten so align:start position:0% deposition of this this tungsten so align:start position:0% deposition of this this tungsten so that's going to be one of the parameters align:start position:0% that's going to be one of the parameters align:start position:0% that's going to be one of the parameters that needs to be empirically align:start position:0% that needs to be empirically align:start position:0% that needs to be empirically studied align:start position:0% studied align:start position:0% studied certainly temperature is going to have a align:start position:0% certainly temperature is going to have a align:start position:0% certainly temperature is going to have a very strong effect align:start position:0% very strong effect align:start position:0% very strong effect in this process on things like uh growth align:start position:0% in this process on things like uh growth align:start position:0% in this process on things like uh growth rate perhaps also align:start position:0% rate perhaps also align:start position:0% rate perhaps also the surface morphology or the roughness align:start position:0% the surface morphology or the roughness align:start position:0% the surface morphology or the roughness of the film the amount of incorporation align:start position:0% of the film the amount of incorporation align:start position:0% of the film the amount of incorporation of tungsten and the ultimate resistivity align:start position:0% of tungsten and the ultimate resistivity align:start position:0% of tungsten and the ultimate resistivity of the film those sorts of parameters align:start position:0% of the film those sorts of parameters align:start position:0% of the film those sorts of parameters and then both hydrogen and tungsten align:start position:0% and then both hydrogen and tungsten align:start position:0% and then both hydrogen and tungsten hexafluoride align:start position:0% hexafluoride align:start position:0% hexafluoride gas flow pressures will also be align:start position:0% align:start position:0% clearly align:start position:0% clearly align:start position:0% clearly of influence in align:start position:0% of influence in align:start position:0% of influence in rate resistivity and so on align:start position:0% rate resistivity and so on align:start position:0% rate resistivity and so on so this is kind of a neat paper because align:start position:0% so this is kind of a neat paper because align:start position:0% so this is kind of a neat paper because they actually are measuring about eight align:start position:0% they actually are measuring about eight align:start position:0% they actually are measuring about eight different parameters align:start position:0% different parameters align:start position:0% different parameters and modeling each of them align:start position:0% and modeling each of them align:start position:0% and modeling each of them so it's kind of a four input eight align:start position:0% so it's kind of a four input eight align:start position:0% so it's kind of a four input eight output uh kind of doe so just to go align:start position:0% output uh kind of doe so just to go align:start position:0% output uh kind of doe so just to go through some of these parameters many of align:start position:0% through some of these parameters many of align:start position:0% through some of these parameters many of them are pretty obvious but a couple of align:start position:0% them are pretty obvious but a couple of align:start position:0% them are pretty obvious but a couple of them i want to explain a little bit more align:start position:0% them i want to explain a little bit more align:start position:0% them i want to explain a little bit more so that you align:start position:0% so that you align:start position:0% so that you have have a good feel for what they are align:start position:0% have have a good feel for what they are align:start position:0% have have a good feel for what they are deposition rate i think that's pretty align:start position:0% deposition rate i think that's pretty align:start position:0% deposition rate i think that's pretty obvious align:start position:0% obvious align:start position:0% obvious generally you would like to have a align:start position:0% generally you would like to have a align:start position:0% generally you would like to have a fairly high rate align:start position:0% fairly high rate align:start position:0% fairly high rate it's going to you know make the process align:start position:0% it's going to you know make the process align:start position:0% it's going to you know make the process time as short as possible align:start position:0% time as short as possible align:start position:0% time as short as possible efficiency in the tool use and so on align:start position:0% efficiency in the tool use and so on align:start position:0% efficiency in the tool use and so on the paper mentions wanting to have about align:start position:0% the paper mentions wanting to have about align:start position:0% the paper mentions wanting to have about 500 nanometers and minute align:start position:0% 500 nanometers and minute align:start position:0% 500 nanometers and minute type type rates align:start position:0% type type rates align:start position:0% type type rates with these processes align:start position:0% with these processes align:start position:0% with these processes the resistivity value align:start position:0% the resistivity value align:start position:0% the resistivity value or row here align:start position:0% or row here align:start position:0% or row here generally you kind of want a low align:start position:0% generally you kind of want a low align:start position:0% generally you kind of want a low resistance or a low resistivity film align:start position:0% resistance or a low resistivity film align:start position:0% resistance or a low resistivity film an interesting twist here is depending align:start position:0% an interesting twist here is depending align:start position:0% an interesting twist here is depending on the application align:start position:0% on the application align:start position:0% on the application that's align:start position:0% that's align:start position:0% that's the actual resistivity you want and align:start position:0% the actual resistivity you want and align:start position:0% the actual resistivity you want and whether it's all that low align:start position:0% whether it's all that low align:start position:0% whether it's all that low actually changes align:start position:0% actually changes align:start position:0% actually changes so for example if you're using it as a align:start position:0% so for example if you're using it as a align:start position:0% so for example if you're using it as a short run metal wire align:start position:0% short run metal wire align:start position:0% short run metal wire pretty much all the time low resistance align:start position:0% pretty much all the time low resistance align:start position:0% pretty much all the time low resistance is important align:start position:0% is important align:start position:0% is important but it turns out interestingly when you align:start position:0% but it turns out interestingly when you align:start position:0% but it turns out interestingly when you use tungsten plugs in conjunction with align:start position:0% use tungsten plugs in conjunction with align:start position:0% use tungsten plugs in conjunction with aluminum layers you may actually want align:start position:0% aluminum layers you may actually want align:start position:0% aluminum layers you may actually want the tungsten plug to be a little bit align:start position:0% the tungsten plug to be a little bit align:start position:0% the tungsten plug to be a little bit higher resistance than the aluminum align:start position:0% higher resistance than the aluminum align:start position:0% higher resistance than the aluminum itself so that because it's a very small align:start position:0% itself so that because it's a very small align:start position:0% itself so that because it's a very small constrained align:start position:0% constrained align:start position:0% constrained spatial area align:start position:0% spatial area align:start position:0% spatial area the higher resistance align:start position:0% the higher resistance align:start position:0% the higher resistance causes the current flow to spread out a align:start position:0% causes the current flow to spread out a align:start position:0% causes the current flow to spread out a little bit more uniformly and actually align:start position:0% little bit more uniformly and actually align:start position:0% little bit more uniformly and actually improves the reliability of these align:start position:0% improves the reliability of these align:start position:0% improves the reliability of these tungsten plugs align:start position:0% tungsten plugs align:start position:0% tungsten plugs so you'd actually like to understand align:start position:0% so you'd actually like to understand align:start position:0% so you'd actually like to understand what the resistance is and be able to align:start position:0% what the resistance is and be able to align:start position:0% what the resistance is and be able to tune that in so that's going to motivate align:start position:0% tune that in so that's going to motivate align:start position:0% tune that in so that's going to motivate some of some of the modeling that that's align:start position:0% some of some of the modeling that that's align:start position:0% some of some of the modeling that that's done here it's not always purely a lower align:start position:0% done here it's not always purely a lower align:start position:0% done here it's not always purely a lower resistance is better align:start position:0% align:start position:0% rs uniformity refers to sheet resistance align:start position:0% rs uniformity refers to sheet resistance align:start position:0% rs uniformity refers to sheet resistance uniformity this is the align:start position:0% uniformity this is the align:start position:0% uniformity this is the ohms per square align:start position:0% ohms per square align:start position:0% ohms per square of of sheet resistance or align:start position:0% of of sheet resistance or align:start position:0% of of sheet resistance or but then a uniformity number calculated align:start position:0% but then a uniformity number calculated align:start position:0% but then a uniformity number calculated for that measured sheet resistance align:start position:0% for that measured sheet resistance align:start position:0% for that measured sheet resistance across the entire wafer and so you want align:start position:0% across the entire wafer and so you want align:start position:0% across the entire wafer and so you want a very good uniformity the three percent align:start position:0% a very good uniformity the three percent align:start position:0% a very good uniformity the three percent non-uniformity number is the deviation align:start position:0% non-uniformity number is the deviation align:start position:0% non-uniformity number is the deviation percentage deviation of the thickness align:start position:0% percentage deviation of the thickness align:start position:0% percentage deviation of the thickness sort of one standard deviation over mean align:start position:0% sort of one standard deviation over mean align:start position:0% sort of one standard deviation over mean is what they're referring to in here i align:start position:0% is what they're referring to in here i align:start position:0% is what they're referring to in here i believe for this uniformity align:start position:0% believe for this uniformity align:start position:0% believe for this uniformity and again one of the motivations there align:start position:0% and again one of the motivations there align:start position:0% and again one of the motivations there is that align:start position:0% is that align:start position:0% is that uniformity in align:start position:0% uniformity in align:start position:0% uniformity in the plasma etch back you want to avoid align:start position:0% the plasma etch back you want to avoid align:start position:0% the plasma etch back you want to avoid recess of these plugs in some region align:start position:0% recess of these plugs in some region align:start position:0% recess of these plugs in some region which can make it difficult for the align:start position:0% which can make it difficult for the align:start position:0% which can make it difficult for the subsequent aluminum to reach down in and align:start position:0% subsequent aluminum to reach down in and align:start position:0% subsequent aluminum to reach down in and actually contact align:start position:0% actually contact align:start position:0% actually contact those recessed plugs align:start position:0% those recessed plugs align:start position:0% those recessed plugs okay align:start position:0% okay align:start position:0% okay uh film stress is another important uh align:start position:0% uh film stress is another important uh align:start position:0% uh film stress is another important uh parameter align:start position:0% parameter align:start position:0% parameter let me see if i've got other slides on align:start position:0% let me see if i've got other slides on align:start position:0% let me see if i've got other slides on that align:start position:0% that align:start position:0% that nope not much align:start position:0% align:start position:0% essentially the uh align:start position:0% essentially the uh align:start position:0% essentially the uh there is a volumetric mismatch to the align:start position:0% there is a volumetric mismatch to the align:start position:0% there is a volumetric mismatch to the deposited tungsten compared to the align:start position:0% deposited tungsten compared to the align:start position:0% deposited tungsten compared to the underlying silicon substrate uh wafer align:start position:0% underlying silicon substrate uh wafer align:start position:0% underlying silicon substrate uh wafer and in fact what you typically find is align:start position:0% and in fact what you typically find is align:start position:0% and in fact what you typically find is that the wafer will undergo align:start position:0% that the wafer will undergo align:start position:0% that the wafer will undergo or be in tensile stress so basically align:start position:0% or be in tensile stress so basically align:start position:0% or be in tensile stress so basically there's more volume in the deposited align:start position:0% there's more volume in the deposited align:start position:0% there's more volume in the deposited film align:start position:0% film align:start position:0% film and it's align:start position:0% and it's align:start position:0% and it's causing align:start position:0% causing align:start position:0% causing a thick a you know a a align:start position:0% a thick a you know a a align:start position:0% a thick a you know a a layer depending on the thickness of of align:start position:0% layer depending on the thickness of of align:start position:0% layer depending on the thickness of of the layer in fact in the in the design align:start position:0% the layer in fact in the in the design align:start position:0% the layer in fact in the in the design they're talking about two different align:start position:0% they're talking about two different align:start position:0% they're talking about two different thicknesses that they explore that align:start position:0% thicknesses that they explore that align:start position:0% thicknesses that they explore that causes the wafer to bow align:start position:0% causes the wafer to bow align:start position:0% causes the wafer to bow so i've got a align:start position:0% so i've got a align:start position:0% so i've got a film that wants to spread out a little align:start position:0% film that wants to spread out a little align:start position:0% film that wants to spread out a little bit more align:start position:0% bit more align:start position:0% bit more especially at lower temperature compared align:start position:0% especially at lower temperature compared align:start position:0% especially at lower temperature compared to the temperature of processing align:start position:0% to the temperature of processing align:start position:0% to the temperature of processing and that wafer bow is a nice indicator align:start position:0% and that wafer bow is a nice indicator align:start position:0% and that wafer bow is a nice indicator of intrinsic internal stress in those align:start position:0% of intrinsic internal stress in those align:start position:0% of intrinsic internal stress in those films align:start position:0% films align:start position:0% films in general you don't want stress align:start position:0% in general you don't want stress align:start position:0% in general you don't want stress mismatches between different layers on align:start position:0% mismatches between different layers on align:start position:0% mismatches between different layers on your wafer align:start position:0% your wafer align:start position:0% your wafer because that means there's sort of this align:start position:0% because that means there's sort of this align:start position:0% because that means there's sort of this excess energy around that would like to align:start position:0% excess energy around that would like to align:start position:0% excess energy around that would like to release itself and one way it releases align:start position:0% release itself and one way it releases align:start position:0% release itself and one way it releases itself is by having those layers align:start position:0% itself is by having those layers align:start position:0% itself is by having those layers delaminate align:start position:0% delaminate align:start position:0% delaminate okay so a reliability concern align:start position:0% okay so a reliability concern align:start position:0% okay so a reliability concern is to try to avoid very high stresses align:start position:0% is to try to avoid very high stresses align:start position:0% is to try to avoid very high stresses in order to support subsequent align:start position:0% in order to support subsequent align:start position:0% in order to support subsequent processing and long-term reliability align:start position:0% processing and long-term reliability align:start position:0% processing and long-term reliability of the material on the wafer align:start position:0% of the material on the wafer align:start position:0% of the material on the wafer okay so they're making some measurements align:start position:0% okay so they're making some measurements align:start position:0% okay so they're making some measurements of film stress in fact the way they make align:start position:0% of film stress in fact the way they make align:start position:0% of film stress in fact the way they make this measurement i believe if i remember align:start position:0% this measurement i believe if i remember align:start position:0% this measurement i believe if i remember right is they measure the wafer bow align:start position:0% right is they measure the wafer bow align:start position:0% right is they measure the wafer bow and then based on align:start position:0% and then based on align:start position:0% and then based on assumed numbers for the modulus align:start position:0% assumed numbers for the modulus align:start position:0% assumed numbers for the modulus the young's modulus align:start position:0% the young's modulus align:start position:0% the young's modulus and perhaps the thickness they may align:start position:0% and perhaps the thickness they may align:start position:0% and perhaps the thickness they may measure the thickness separately align:start position:0% measure the thickness separately align:start position:0% measure the thickness separately they basically calculate or or estimate align:start position:0% they basically calculate or or estimate align:start position:0% they basically calculate or or estimate what the film stress stress align:start position:0% align:start position:0% in some sense at whatever temperature align:start position:0% in some sense at whatever temperature align:start position:0% in some sense at whatever temperature you're you're doing if you could deposit align:start position:0% you're you're doing if you could deposit align:start position:0% you're you're doing if you could deposit the film and keep the wafer at 300 align:start position:0% the film and keep the wafer at 300 align:start position:0% the film and keep the wafer at 300 degrees or 100 whatever it is align:start position:0% degrees or 100 whatever it is align:start position:0% degrees or 100 whatever it is at that point in time i don't believe align:start position:0% at that point in time i don't believe align:start position:0% at that point in time i don't believe the wafer undergoes or experiences much align:start position:0% the wafer undergoes or experiences much align:start position:0% the wafer undergoes or experiences much stress during this deposition but then align:start position:0% stress during this deposition but then align:start position:0% stress during this deposition but then it's a align:start position:0% it's a align:start position:0% it's a temperature expansion difference align:start position:0% temperature expansion difference align:start position:0% temperature expansion difference i'm not entirely sure of that it's align:start position:0% i'm not entirely sure of that it's align:start position:0% i'm not entirely sure of that it's possible that the simple incorporation align:start position:0% possible that the simple incorporation align:start position:0% possible that the simple incorporation at the time align:start position:0% at the time align:start position:0% at the time of deposition align:start position:0% of deposition align:start position:0% of deposition also has some stress but really we're align:start position:0% also has some stress but really we're align:start position:0% also has some stress but really we're talking about intrinsic stress that our align:start position:0% talking about intrinsic stress that our align:start position:0% talking about intrinsic stress that our extrinsic stress that align:start position:0% extrinsic stress that align:start position:0% extrinsic stress that remains at the end of into processing align:start position:0% align:start position:0% know any more on that hayden align:start position:0% align:start position:0% i mean there are there are situations align:start position:0% i mean there are there are situations align:start position:0% i mean there are there are situations where align:start position:0% where align:start position:0% where you're depositing align:start position:0% you're depositing align:start position:0% you're depositing you know an epitaxial layer of say a align:start position:0% you know an epitaxial layer of say a align:start position:0% you know an epitaxial layer of say a semiconductor align:start position:0% semiconductor align:start position:0% semiconductor where that has a very clearly defined align:start position:0% where that has a very clearly defined align:start position:0% where that has a very clearly defined lattice whose pitch is mismatched with align:start position:0% lattice whose pitch is mismatched with align:start position:0% lattice whose pitch is mismatched with the substrate underneath in that case align:start position:0% the substrate underneath in that case align:start position:0% the substrate underneath in that case you know there's i guess a definite align:start position:0% you know there's i guess a definite align:start position:0% you know there's i guess a definite uh align:start position:0% uh align:start position:0% uh definite built-in stress that's going to align:start position:0% definite built-in stress that's going to align:start position:0% definite built-in stress that's going to result in that deposition that isn't align:start position:0% result in that deposition that isn't align:start position:0% result in that deposition that isn't connected with thermal contraction but i align:start position:0% connected with thermal contraction but i align:start position:0% connected with thermal contraction but i would have thought align:start position:0% would have thought align:start position:0% would have thought in this situation align:start position:0% in this situation align:start position:0% in this situation it's mostly align:start position:0% it's mostly align:start position:0% it's mostly it's mostly to do with align:start position:0% it's mostly to do with align:start position:0% it's mostly to do with you know the presence of impurities align:start position:0% you know the presence of impurities align:start position:0% you know the presence of impurities changing the cte of the deposited life align:start position:0% changing the cte of the deposited life align:start position:0% changing the cte of the deposited life that align:start position:0% that align:start position:0% that stretch align:start position:0% align:start position:0% good align:start position:0% align:start position:0% the fifth parameter that they measure align:start position:0% the fifth parameter that they measure align:start position:0% the fifth parameter that they measure and model is step coverage and so here align:start position:0% and model is step coverage and so here align:start position:0% and model is step coverage and so here that refers to align:start position:0% that refers to align:start position:0% that refers to essentially how good the deposition align:start position:0% essentially how good the deposition align:start position:0% essentially how good the deposition process is align:start position:0% process is align:start position:0% process is right here align:start position:0% right here align:start position:0% right here in filling align:start position:0% in filling align:start position:0% in filling filling these features now if all of align:start position:0% filling these features now if all of align:start position:0% filling these features now if all of your features are exactly the same size align:start position:0% your features are exactly the same size align:start position:0% your features are exactly the same size then that's that's a little bit easier align:start position:0% then that's that's a little bit easier align:start position:0% then that's that's a little bit easier so if all you're doing is filling align:start position:0% so if all you're doing is filling align:start position:0% so if all you're doing is filling contacts of exactly the same size align:start position:0% contacts of exactly the same size align:start position:0% contacts of exactly the same size then you've got a nice well-defined align:start position:0% then you've got a nice well-defined align:start position:0% then you've got a nice well-defined number for step coverage and that's align:start position:0% number for step coverage and that's align:start position:0% number for step coverage and that's really what they're doing in this align:start position:0% really what they're doing in this align:start position:0% really what they're doing in this application but i should mention that align:start position:0% application but i should mention that align:start position:0% application but i should mention that these kind of conformal depositions can align:start position:0% these kind of conformal depositions can align:start position:0% these kind of conformal depositions can actually be a very strong challenge for align:start position:0% actually be a very strong challenge for align:start position:0% actually be a very strong challenge for dealing with patterns or features of align:start position:0% dealing with patterns or features of align:start position:0% dealing with patterns or features of different sizes so you can imagine for align:start position:0% different sizes so you can imagine for align:start position:0% different sizes so you can imagine for example align:start position:0% example align:start position:0% example having some contacts here align:start position:0% having some contacts here align:start position:0% having some contacts here having some structures that might in align:start position:0% having some structures that might in align:start position:0% having some structures that might in fact be lines then having some other align:start position:0% fact be lines then having some other align:start position:0% fact be lines then having some other structures that are in fact very wide align:start position:0% structures that are in fact very wide align:start position:0% structures that are in fact very wide lines align:start position:0% lines align:start position:0% lines and now if i can formally fill i might align:start position:0% and now if i can formally fill i might align:start position:0% and now if i can formally fill i might be able to fill the narrow structures align:start position:0% align:start position:0% but the wider ones will not necessarily align:start position:0% but the wider ones will not necessarily align:start position:0% but the wider ones will not necessarily fill completely i won't have excellent align:start position:0% fill completely i won't have excellent align:start position:0% fill completely i won't have excellent excellent step coverage now i believe align:start position:0% excellent step coverage now i believe align:start position:0% excellent step coverage now i believe what they're doing in this align:start position:0% what they're doing in this align:start position:0% what they're doing in this in this align:start position:0% in this align:start position:0% in this paper align:start position:0% paper align:start position:0% paper is really just looking locally align:start position:0% is really just looking locally align:start position:0% is really just looking locally after the deposition and looking to see align:start position:0% after the deposition and looking to see align:start position:0% after the deposition and looking to see sort of what percentage of volume align:start position:0% sort of what percentage of volume align:start position:0% sort of what percentage of volume is missing in that little in that little align:start position:0% is missing in that little in that little align:start position:0% is missing in that little in that little uh indentation resulting from the align:start position:0% uh indentation resulting from the align:start position:0% uh indentation resulting from the conformal deposition and looking for align:start position:0% conformal deposition and looking for align:start position:0% conformal deposition and looking for fairly high near 100 align:start position:0% fairly high near 100 align:start position:0% fairly high near 100 step coverage so that there's not a align:start position:0% step coverage so that there's not a align:start position:0% step coverage so that there's not a missing percentage align:start position:0% missing percentage align:start position:0% missing percentage of the feature align:start position:0% of the feature align:start position:0% of the feature compared to say align:start position:0% compared to say align:start position:0% compared to say one of these large structures that might align:start position:0% one of these large structures that might align:start position:0% one of these large structures that might have a very high missing missing volume align:start position:0% have a very high missing missing volume align:start position:0% have a very high missing missing volume generally they're able to achieve fairly align:start position:0% generally they're able to achieve fairly align:start position:0% generally they're able to achieve fairly high align:start position:0% high align:start position:0% high step coverages so that's not one of the align:start position:0% step coverages so that's not one of the align:start position:0% step coverages so that's not one of the critical critical output parameters in align:start position:0% critical critical output parameters in align:start position:0% critical critical output parameters in this process align:start position:0% this process align:start position:0% this process now also shown on this this picture and align:start position:0% now also shown on this this picture and align:start position:0% now also shown on this this picture and maybe a little bit hard for you to see align:start position:0% maybe a little bit hard for you to see align:start position:0% maybe a little bit hard for you to see it's also in the paper but on the right align:start position:0% it's also in the paper but on the right align:start position:0% it's also in the paper but on the right you see a align:start position:0% you see a align:start position:0% you see a top-down photograph or or sem micrograph align:start position:0% top-down photograph or or sem micrograph align:start position:0% top-down photograph or or sem micrograph of the surface structure and it gives align:start position:0% of the surface structure and it gives align:start position:0% of the surface structure and it gives you a sense a little bit of two things align:start position:0% you a sense a little bit of two things align:start position:0% you a sense a little bit of two things one is there's a lot of surface align:start position:0% one is there's a lot of surface align:start position:0% one is there's a lot of surface morphology going on align:start position:0% morphology going on align:start position:0% morphology going on surface roughness but the other basic align:start position:0% surface roughness but the other basic align:start position:0% surface roughness but the other basic thing that happens is as you're align:start position:0% thing that happens is as you're align:start position:0% thing that happens is as you're depositing these uh align:start position:0% depositing these uh align:start position:0% depositing these uh these materials they are lining up in align:start position:0% these materials they are lining up in align:start position:0% these materials they are lining up in sort of a polycrystalline kind of kind align:start position:0% sort of a polycrystalline kind of kind align:start position:0% sort of a polycrystalline kind of kind of way you do get some local align:start position:0% of way you do get some local align:start position:0% of way you do get some local self-organization you can see that also align:start position:0% self-organization you can see that also align:start position:0% self-organization you can see that also over here on the on the left side if you align:start position:0% over here on the on the left side if you align:start position:0% over here on the on the left side if you kind of see there's little crystals or align:start position:0% kind of see there's little crystals or align:start position:0% kind of see there's little crystals or little threads or align:start position:0% little threads or align:start position:0% little threads or and in here you can see align:start position:0% and in here you can see align:start position:0% and in here you can see see a little bit of the polycrystalline align:start position:0% see a little bit of the polycrystalline align:start position:0% see a little bit of the polycrystalline structure and that's actually one of the align:start position:0% structure and that's actually one of the align:start position:0% structure and that's actually one of the things that gives that surface roughness align:start position:0% things that gives that surface roughness align:start position:0% things that gives that surface roughness it's not surface roughness on the atom align:start position:0% it's not surface roughness on the atom align:start position:0% it's not surface roughness on the atom by atom level it's really these align:start position:0% by atom level it's really these align:start position:0% by atom level it's really these preferential growth directions in some align:start position:0% preferential growth directions in some align:start position:0% preferential growth directions in some of these some of these grains if you align:start position:0% of these some of these grains if you align:start position:0% of these some of these grains if you will align:start position:0% will align:start position:0% will as as they're growing align:start position:0% as as they're growing align:start position:0% as as they're growing okay align:start position:0% align:start position:0% uh align:start position:0% uh align:start position:0% uh only two or three more parameters here align:start position:0% only two or three more parameters here align:start position:0% only two or three more parameters here another one and this is kind of neat i align:start position:0% another one and this is kind of neat i align:start position:0% another one and this is kind of neat i think because it's a align:start position:0% think because it's a align:start position:0% think because it's a nice parameter that contrasts with the align:start position:0% nice parameter that contrasts with the align:start position:0% nice parameter that contrasts with the other ones the other ones are let's see align:start position:0% other ones the other ones are let's see align:start position:0% other ones the other ones are let's see most of these things like step coverage align:start position:0% most of these things like step coverage align:start position:0% most of these things like step coverage film stress uniformity resistance align:start position:0% film stress uniformity resistance align:start position:0% film stress uniformity resistance uniformity those are parameters of the align:start position:0% uniformity those are parameters of the align:start position:0% uniformity those are parameters of the final wafer or the intermediate wafer align:start position:0% final wafer or the intermediate wafer align:start position:0% final wafer or the intermediate wafer state that you want to achieve align:start position:0% state that you want to achieve align:start position:0% state that you want to achieve deposition rate and wf6 conversion are align:start position:0% deposition rate and wf6 conversion are align:start position:0% deposition rate and wf6 conversion are both more manufacturing parameters align:start position:0% both more manufacturing parameters align:start position:0% both more manufacturing parameters in some sense they don't really matter align:start position:0% in some sense they don't really matter align:start position:0% in some sense they don't really matter to the final wafer product but they can align:start position:0% to the final wafer product but they can align:start position:0% to the final wafer product but they can matter a great deal for the cost align:start position:0% matter a great deal for the cost align:start position:0% matter a great deal for the cost efficiency or the time efficiency of the align:start position:0% efficiency or the time efficiency of the align:start position:0% efficiency or the time efficiency of the process align:start position:0% process align:start position:0% process so wf6 conversion is what percentage of align:start position:0% so wf6 conversion is what percentage of align:start position:0% so wf6 conversion is what percentage of the incoming align:start position:0% the incoming align:start position:0% the incoming gas is actually resulting in align:start position:0% gas is actually resulting in align:start position:0% gas is actually resulting in the deposited film so you might have align:start position:0% the deposited film so you might have align:start position:0% the deposited film so you might have uh only you know one percent of the align:start position:0% uh only you know one percent of the align:start position:0% uh only you know one percent of the tungsten align:start position:0% align:start position:0% going on to the wafer and 99 percent align:start position:0% going on to the wafer and 99 percent align:start position:0% going on to the wafer and 99 percent going out the exhaust pipe align:start position:0% going out the exhaust pipe align:start position:0% going out the exhaust pipe that's terrible waste they have some align:start position:0% that's terrible waste they have some align:start position:0% that's terrible waste they have some environmental implications align:start position:0% environmental implications align:start position:0% environmental implications but it also uh has a lot of cost align:start position:0% but it also uh has a lot of cost align:start position:0% but it also uh has a lot of cost implications because tungsten align:start position:0% implications because tungsten align:start position:0% implications because tungsten hexafluoride is not a cheap gas align:start position:0% hexafluoride is not a cheap gas align:start position:0% hexafluoride is not a cheap gas you know very very highly purified so i align:start position:0% you know very very highly purified so i align:start position:0% you know very very highly purified so i like this paper in terms of they're align:start position:0% like this paper in terms of they're align:start position:0% like this paper in terms of they're actually trying to track and keep track align:start position:0% actually trying to track and keep track align:start position:0% actually trying to track and keep track of align:start position:0% of align:start position:0% of and model align:start position:0% and model align:start position:0% and model the ws6 conversion rate and try to align:start position:0% the ws6 conversion rate and try to align:start position:0% the ws6 conversion rate and try to optimize that align:start position:0% optimize that align:start position:0% optimize that parameter also align:start position:0% parameter also align:start position:0% parameter also then finally a couple more align:start position:0% then finally a couple more align:start position:0% then finally a couple more parameters here reflectance is basically align:start position:0% parameters here reflectance is basically align:start position:0% parameters here reflectance is basically an optical reflectance uh percentage align:start position:0% an optical reflectance uh percentage align:start position:0% an optical reflectance uh percentage light reflected uh kind of measurement align:start position:0% light reflected uh kind of measurement align:start position:0% light reflected uh kind of measurement uh of the surface align:start position:0% uh of the surface align:start position:0% uh of the surface topography so that's how they're trying align:start position:0% topography so that's how they're trying align:start position:0% topography so that's how they're trying to estimate or correlate to the align:start position:0% to estimate or correlate to the align:start position:0% to estimate or correlate to the roughness of the surface film align:start position:0% roughness of the surface film align:start position:0% roughness of the surface film and then finally another manufacturing align:start position:0% and then finally another manufacturing align:start position:0% and then finally another manufacturing parameter is align:start position:0% parameter is align:start position:0% parameter is reproducibility not just wafer align:start position:0% reproducibility not just wafer align:start position:0% reproducibility not just wafer uniformity align:start position:0% uniformity align:start position:0% uniformity you know wait for one one wafer at a align:start position:0% you know wait for one one wafer at a align:start position:0% you know wait for one one wafer at a time but wait for it to wait for align:start position:0% time but wait for it to wait for align:start position:0% time but wait for it to wait for repeatability or reproducibility of the align:start position:0% repeatability or reproducibility of the align:start position:0% repeatability or reproducibility of the process so they are actually making in align:start position:0% process so they are actually making in align:start position:0% process so they are actually making in this case i think the assumption align:start position:0% this case i think the assumption align:start position:0% this case i think the assumption that pretty much all of these processes align:start position:0% that pretty much all of these processes align:start position:0% that pretty much all of these processes are align:start position:0% align:start position:0% more or less similarly repeatable but align:start position:0% more or less similarly repeatable but align:start position:0% more or less similarly repeatable but they want to estimate how align:start position:0% they want to estimate how align:start position:0% they want to estimate how repeatable the process is and we'll see align:start position:0% repeatable the process is and we'll see align:start position:0% repeatable the process is and we'll see that in their doe they're going to align:start position:0% that in their doe they're going to align:start position:0% that in their doe they're going to basically do some replicates align:start position:0% basically do some replicates align:start position:0% basically do some replicates at one of the design points and use that align:start position:0% at one of the design points and use that align:start position:0% at one of the design points and use that to estimate reproducibility align:start position:0% to estimate reproducibility align:start position:0% to estimate reproducibility so they're not going in and trying to align:start position:0% so they're not going in and trying to align:start position:0% so they're not going in and trying to estimate where align:start position:0% estimate where align:start position:0% estimate where the noise or the variance is minimum align:start position:0% the noise or the variance is minimum align:start position:0% the noise or the variance is minimum across their operating space they're align:start position:0% across their operating space they're align:start position:0% across their operating space they're actually assuming that that's align:start position:0% actually assuming that that's align:start position:0% actually assuming that that's that align:start position:0% that align:start position:0% that similar okay align:start position:0% align:start position:0% so align:start position:0% so align:start position:0% so here's what they do in the paper their align:start position:0% here's what they do in the paper their align:start position:0% here's what they do in the paper their goals are to build response service align:start position:0% goals are to build response service align:start position:0% goals are to build response service models just classic response surface align:start position:0% models just classic response surface align:start position:0% models just classic response surface models align:start position:0% models align:start position:0% models and then i as i said the reproducibility align:start position:0% and then i as i said the reproducibility align:start position:0% and then i as i said the reproducibility is really focused just on the center align:start position:0% is really focused just on the center align:start position:0% is really focused just on the center point designs align:start position:0% point designs align:start position:0% point designs and applying that throughout the entire align:start position:0% and applying that throughout the entire align:start position:0% and applying that throughout the entire space align:start position:0% space align:start position:0% space and then they want to use these response align:start position:0% and then they want to use these response align:start position:0% and then they want to use these response surface models align:start position:0% surface models align:start position:0% surface models they want to qualitatively understand align:start position:0% they want to qualitatively understand align:start position:0% they want to qualitatively understand some of the trends align:start position:0% some of the trends align:start position:0% some of the trends very good engineering use you know align:start position:0% very good engineering use you know align:start position:0% very good engineering use you know trying to understand from an engineering align:start position:0% trying to understand from an engineering align:start position:0% trying to understand from an engineering point and then they also want to use align:start position:0% point and then they also want to use align:start position:0% point and then they also want to use these models in an optimization align:start position:0% these models in an optimization align:start position:0% these models in an optimization kind of loop to try to drive towards align:start position:0% kind of loop to try to drive towards align:start position:0% kind of loop to try to drive towards satisfaction of multiple align:start position:0% satisfaction of multiple align:start position:0% satisfaction of multiple limits or goals on those align:start position:0% limits or goals on those align:start position:0% limits or goals on those what was it seven or eight uh output align:start position:0% what was it seven or eight uh output align:start position:0% what was it seven or eight uh output parameters align:start position:0% parameters align:start position:0% parameters okay align:start position:0% okay align:start position:0% okay now the paper does refer to some prior align:start position:0% now the paper does refer to some prior align:start position:0% now the paper does refer to some prior work on screening experiments in fact align:start position:0% work on screening experiments in fact align:start position:0% work on screening experiments in fact uh these uh align:start position:0% uh these uh align:start position:0% uh these uh precision 5000 tools and most fab tools align:start position:0% precision 5000 tools and most fab tools align:start position:0% precision 5000 tools and most fab tools probably have align:start position:0% probably have align:start position:0% probably have literally in some cases dozens of align:start position:0% literally in some cases dozens of align:start position:0% literally in some cases dozens of control settings some align:start position:0% control settings some align:start position:0% control settings some subset of which maybe only align:start position:0% subset of which maybe only align:start position:0% subset of which maybe only align:start position:0% align:start position:0% are easily settable electronically and align:start position:0% are easily settable electronically and align:start position:0% are easily settable electronically and another 15 might be more mechanical or align:start position:0% another 15 might be more mechanical or align:start position:0% another 15 might be more mechanical or other you know more permanent align:start position:0% other you know more permanent align:start position:0% other you know more permanent adjustments but there's a huge space of align:start position:0% adjustments but there's a huge space of align:start position:0% adjustments but there's a huge space of possible parameters the paper just align:start position:0% possible parameters the paper just align:start position:0% possible parameters the paper just refers to prior screening experiments to align:start position:0% refers to prior screening experiments to align:start position:0% refers to prior screening experiments to narrow down to these four parameters so align:start position:0% narrow down to these four parameters so align:start position:0% narrow down to these four parameters so i don't want you to think that they knew align:start position:0% i don't want you to think that they knew align:start position:0% i don't want you to think that they knew automatically up front what four align:start position:0% automatically up front what four align:start position:0% automatically up front what four parameters they wanted to play with align:start position:0% align:start position:0% there is a reference if you want to go align:start position:0% there is a reference if you want to go align:start position:0% there is a reference if you want to go try to track it down further for their align:start position:0% try to track it down further for their align:start position:0% try to track it down further for their prior work it's actually in some obscure align:start position:0% prior work it's actually in some obscure align:start position:0% prior work it's actually in some obscure conference thing so it's hard to hard to align:start position:0% conference thing so it's hard to hard to align:start position:0% conference thing so it's hard to hard to find um align:start position:0% find um align:start position:0% find um so the details exactly of what they did align:start position:0% so the details exactly of what they did align:start position:0% so the details exactly of what they did in their in their screening are not align:start position:0% in their in their screening are not align:start position:0% in their in their screening are not entirely clear but the point is they align:start position:0% entirely clear but the point is they align:start position:0% entirely clear but the point is they probably did something like align:start position:0% probably did something like align:start position:0% probably did something like a fractional factorial align:start position:0% a fractional factorial align:start position:0% a fractional factorial highly fractionalized factorial kind of align:start position:0% highly fractionalized factorial kind of align:start position:0% highly fractionalized factorial kind of design align:start position:0% design align:start position:0% design to explore the space of just trying to align:start position:0% to explore the space of just trying to align:start position:0% to explore the space of just trying to see what parameters are significant what align:start position:0% see what parameters are significant what align:start position:0% see what parameters are significant what influence what input parameters align:start position:0% influence what input parameters align:start position:0% influence what input parameters influence the outputs that they were align:start position:0% influence the outputs that they were align:start position:0% influence the outputs that they were of concern and then narrow that down to align:start position:0% of concern and then narrow that down to align:start position:0% of concern and then narrow that down to just these four parameters that they align:start position:0% just these four parameters that they align:start position:0% just these four parameters that they wanted to do align:start position:0% wanted to do align:start position:0% wanted to do a second set of experiments which is align:start position:0% a second set of experiments which is align:start position:0% a second set of experiments which is what's in this in this paper that are align:start position:0% what's in this in this paper that are align:start position:0% what's in this in this paper that are focused on response surface modeling align:start position:0% focused on response surface modeling align:start position:0% focused on response surface modeling okay align:start position:0% okay align:start position:0% okay there's another align:start position:0% there's another align:start position:0% there's another interesting subtle point that align:start position:0% interesting subtle point that align:start position:0% interesting subtle point that pops up in the in the description of the align:start position:0% pops up in the in the description of the align:start position:0% pops up in the in the description of the paper align:start position:0% paper align:start position:0% paper some of these parameters things like align:start position:0% some of these parameters things like align:start position:0% some of these parameters things like sheet resistance sheet resistance is align:start position:0% sheet resistance sheet resistance is align:start position:0% sheet resistance sheet resistance is resistivity align:start position:0% resistivity align:start position:0% resistivity uh align:start position:0% align:start position:0% is it divided by align:start position:0% is it divided by align:start position:0% is it divided by per square resistivity the the pure align:start position:0% per square resistivity the the pure align:start position:0% per square resistivity the the pure resistivity divided by the thickness align:start position:0% resistivity divided by the thickness align:start position:0% resistivity divided by the thickness uh of the film align:start position:0% uh of the film align:start position:0% uh of the film okay align:start position:0% okay align:start position:0% okay so that the thickness of the deposited align:start position:0% so that the thickness of the deposited align:start position:0% so that the thickness of the deposited film will change the rs measure align:start position:0% film will change the rs measure align:start position:0% film will change the rs measure on the wafer okay align:start position:0% on the wafer okay align:start position:0% on the wafer okay what that means is to do a fair align:start position:0% what that means is to do a fair align:start position:0% what that means is to do a fair comparison just of truly the resistivity align:start position:0% comparison just of truly the resistivity align:start position:0% comparison just of truly the resistivity of the film align:start position:0% of the film align:start position:0% of the film and uh or to estimate that you actually align:start position:0% and uh or to estimate that you actually align:start position:0% and uh or to estimate that you actually want to grow align:start position:0% want to grow align:start position:0% want to grow all of your films to about the same align:start position:0% all of your films to about the same align:start position:0% all of your films to about the same thickness align:start position:0% thickness align:start position:0% thickness well if the deposition rate is a align:start position:0% well if the deposition rate is a align:start position:0% well if the deposition rate is a parameter that you don't necessarily align:start position:0% parameter that you don't necessarily align:start position:0% parameter that you don't necessarily know and you're trying to model how do align:start position:0% know and you're trying to model how do align:start position:0% know and you're trying to model how do you do that align:start position:0% align:start position:0% so in other words if i run for a fixed align:start position:0% so in other words if i run for a fixed align:start position:0% so in other words if i run for a fixed period of time maybe i run for 60 align:start position:0% period of time maybe i run for 60 align:start position:0% period of time maybe i run for 60 seconds on each of these design points align:start position:0% seconds on each of these design points align:start position:0% seconds on each of these design points but i have deposition rates that are align:start position:0% but i have deposition rates that are align:start position:0% but i have deposition rates that are different by a factor of 2 5 10 i'm align:start position:0% different by a factor of 2 5 10 i'm align:start position:0% different by a factor of 2 5 10 i'm going to get widely varying thicknesses align:start position:0% going to get widely varying thicknesses align:start position:0% going to get widely varying thicknesses and therefore when i make some of these align:start position:0% and therefore when i make some of these align:start position:0% and therefore when i make some of these simple sheet resistance measurements align:start position:0% simple sheet resistance measurements align:start position:0% simple sheet resistance measurements that's that's going to be very very align:start position:0% that's that's going to be very very align:start position:0% that's that's going to be very very tricky align:start position:0% tricky align:start position:0% tricky furthermore if if my thickness is align:start position:0% furthermore if if my thickness is align:start position:0% furthermore if if my thickness is varying align:start position:0% varying align:start position:0% varying that makes it basis for comparison of align:start position:0% that makes it basis for comparison of align:start position:0% that makes it basis for comparison of many of these other parameters really align:start position:0% many of these other parameters really align:start position:0% many of these other parameters really hard align:start position:0% hard align:start position:0% hard how about film stress align:start position:0% how about film stress align:start position:0% how about film stress well is it a film stress that i'm seeing align:start position:0% well is it a film stress that i'm seeing align:start position:0% well is it a film stress that i'm seeing because of the thickness difference or align:start position:0% because of the thickness difference or align:start position:0% because of the thickness difference or is it truly intrinsic film stress in the align:start position:0% is it truly intrinsic film stress in the align:start position:0% is it truly intrinsic film stress in the in align:start position:0% in align:start position:0% in the layer align:start position:0% the layer align:start position:0% the layer so what they've done is been and i think align:start position:0% so what they've done is been and i think align:start position:0% so what they've done is been and i think this is very cool they were very careful align:start position:0% this is very cool they were very careful align:start position:0% this is very cool they were very careful and they align:start position:0% and they align:start position:0% and they say that they are doing this to where align:start position:0% say that they are doing this to where align:start position:0% say that they are doing this to where all films are grown to about one micron align:start position:0% all films are grown to about one micron align:start position:0% all films are grown to about one micron thickness in this align:start position:0% thickness in this align:start position:0% thickness in this doe what's sneaky is i think to actually align:start position:0% doe what's sneaky is i think to actually align:start position:0% doe what's sneaky is i think to actually do this align:start position:0% do this align:start position:0% do this which is also good because that align:start position:0% which is also good because that align:start position:0% which is also good because that is really the the thickness that's close align:start position:0% is really the the thickness that's close align:start position:0% is really the the thickness that's close to their target application align:start position:0% to their target application align:start position:0% to their target application what's what's a little bit sneaky is to align:start position:0% what's what's a little bit sneaky is to align:start position:0% what's what's a little bit sneaky is to do that i think they basically had to do align:start position:0% do that i think they basically had to do align:start position:0% do that i think they basically had to do prior runs maybe at each individual voe align:start position:0% prior runs maybe at each individual voe align:start position:0% prior runs maybe at each individual voe point align:start position:0% point align:start position:0% point to align:start position:0% to align:start position:0% to do the run align:start position:0% do the run align:start position:0% do the run measure the thickness align:start position:0% measure the thickness align:start position:0% measure the thickness estimate the growth rate align:start position:0% estimate the growth rate align:start position:0% estimate the growth rate and then adjust the time for the real align:start position:0% and then adjust the time for the real align:start position:0% and then adjust the time for the real doe run align:start position:0% doe run align:start position:0% doe run on each of the design points so i think align:start position:0% on each of the design points so i think align:start position:0% on each of the design points so i think that there's a whole extra duplicate set align:start position:0% that there's a whole extra duplicate set align:start position:0% that there's a whole extra duplicate set of experiments that they're not talking align:start position:0% of experiments that they're not talking align:start position:0% of experiments that they're not talking about align:start position:0% about align:start position:0% about that that that were done in order to be align:start position:0% that that that were done in order to be align:start position:0% that that that were done in order to be able to normalize align:start position:0% able to normalize align:start position:0% able to normalize and get a very nice align:start position:0% and get a very nice align:start position:0% and get a very nice set of experiments in here align:start position:0% set of experiments in here align:start position:0% set of experiments in here that they could talk about and be able align:start position:0% that they could talk about and be able align:start position:0% that they could talk about and be able to use to use align:start position:0% to use to use align:start position:0% to use to use to make the right kinds of optimization align:start position:0% to make the right kinds of optimization align:start position:0% to make the right kinds of optimization and judgment align:start position:0% and judgment align:start position:0% and judgment so it's kind of an interesting lesson i align:start position:0% so it's kind of an interesting lesson i align:start position:0% so it's kind of an interesting lesson i think there may often be additional align:start position:0% think there may often be additional align:start position:0% think there may often be additional experimentation around align:start position:0% experimentation around align:start position:0% experimentation around your main doe that you have to do in align:start position:0% your main doe that you have to do in align:start position:0% your main doe that you have to do in order for your main doe to really be a align:start position:0% order for your main doe to really be a align:start position:0% order for your main doe to really be a good clean align:start position:0% good clean align:start position:0% good clean doe align:start position:0% doe align:start position:0% doe now it's possible that maybe align:start position:0% now it's possible that maybe align:start position:0% now it's possible that maybe their screening experiment gave them align:start position:0% their screening experiment gave them align:start position:0% their screening experiment gave them some information to be able to do a align:start position:0% some information to be able to do a align:start position:0% some information to be able to do a rough estimate a good enough estimate of align:start position:0% rough estimate a good enough estimate of align:start position:0% rough estimate a good enough estimate of deposition rate so it's possible they align:start position:0% deposition rate so it's possible they align:start position:0% deposition rate so it's possible they were able to get that benefit from from align:start position:0% were able to get that benefit from from align:start position:0% were able to get that benefit from from their prior screening doe align:start position:0% align:start position:0% okay here's their experimental design align:start position:0% okay here's their experimental design align:start position:0% okay here's their experimental design what they're going to do is a central align:start position:0% what they're going to do is a central align:start position:0% what they're going to do is a central composite with five levels align:start position:0% composite with five levels align:start position:0% composite with five levels at each align:start position:0% at each align:start position:0% at each design point which is pretty cool i align:start position:0% design point which is pretty cool i align:start position:0% design point which is pretty cool i think that's better than the three align:start position:0% think that's better than the three align:start position:0% think that's better than the three levels align:start position:0% levels align:start position:0% levels what's one reason that you might like align:start position:0% what's one reason that you might like align:start position:0% what's one reason that you might like five levels align:start position:0% five levels align:start position:0% five levels rather than three especially if you want align:start position:0% rather than three especially if you want align:start position:0% rather than three especially if you want to build say quadratic response surface align:start position:0% to build say quadratic response surface align:start position:0% to build say quadratic response surface models align:start position:0% align:start position:0% we haven't explicitly done this before align:start position:0% we haven't explicitly done this before align:start position:0% we haven't explicitly done this before but align:start position:0% but align:start position:0% but for example on on on quiz 2 align:start position:0% for example on on on quiz 2 align:start position:0% for example on on on quiz 2 you're going from a linear model to a align:start position:0% you're going from a linear model to a align:start position:0% you're going from a linear model to a quadratic model and asking questions align:start position:0% quadratic model and asking questions align:start position:0% quadratic model and asking questions about lack of fit align:start position:0% about lack of fit align:start position:0% about lack of fit in order to say whether a linear model align:start position:0% align:start position:0% had lack of fit i needed a third design align:start position:0% had lack of fit i needed a third design align:start position:0% had lack of fit i needed a third design point align:start position:0% point align:start position:0% point right so that i could say well align:start position:0% right so that i could say well align:start position:0% right so that i could say well does that third design point give me align:start position:0% does that third design point give me align:start position:0% does that third design point give me indication that the linear model was not align:start position:0% indication that the linear model was not align:start position:0% indication that the linear model was not sufficient if i do three levels align:start position:0% sufficient if i do three levels align:start position:0% sufficient if i do three levels and build a quadratic model align:start position:0% and build a quadratic model align:start position:0% and build a quadratic model and i want to ask the question is a align:start position:0% and i want to ask the question is a align:start position:0% and i want to ask the question is a quadratic model sufficient align:start position:0% quadratic model sufficient align:start position:0% quadratic model sufficient or is there lack of fit in the quadratic align:start position:0% or is there lack of fit in the quadratic align:start position:0% or is there lack of fit in the quadratic model align:start position:0% model align:start position:0% model can i do that align:start position:0% can i do that align:start position:0% can i do that with just three align:start position:0% with just three align:start position:0% with just three data points align:start position:0% align:start position:0% no right you can fit a quadratic two to align:start position:0% no right you can fit a quadratic two to align:start position:0% no right you can fit a quadratic two to three data points so align:start position:0% align:start position:0% you won't detect if there's some other align:start position:0% you won't detect if there's some other align:start position:0% you won't detect if there's some other structure going on so in order to be align:start position:0% structure going on so in order to be align:start position:0% structure going on so in order to be able to do a lack of fit analysis up to align:start position:0% able to do a lack of fit analysis up to align:start position:0% able to do a lack of fit analysis up to seeing if a quadratic model exhibits align:start position:0% seeing if a quadratic model exhibits align:start position:0% seeing if a quadratic model exhibits then you need at least additional levels align:start position:0% then you need at least additional levels align:start position:0% then you need at least additional levels implicitly somewhere in the design so align:start position:0% implicitly somewhere in the design so align:start position:0% implicitly somewhere in the design so that's kind of a neat a neat thing that align:start position:0% that's kind of a neat a neat thing that align:start position:0% that's kind of a neat a neat thing that they do here align:start position:0% they do here align:start position:0% they do here so what they've done is normalized all align:start position:0% so what they've done is normalized all align:start position:0% so what they've done is normalized all of their factors their five five four align:start position:0% of their factors their five five four align:start position:0% of their factors their five five four factors here space temperature and so on align:start position:0% factors here space temperature and so on align:start position:0% factors here space temperature and so on again not just two in our in their case align:start position:0% again not just two in our in their case align:start position:0% again not just two in our in their case not just to a minus one plus one but uh align:start position:0% not just to a minus one plus one but uh align:start position:0% not just to a minus one plus one but uh they use with their five levels a minus align:start position:0% they use with their five levels a minus align:start position:0% they use with their five levels a minus two to plus two normalization but align:start position:0% two to plus two normalization but align:start position:0% two to plus two normalization but essentially they're doing five up to align:start position:0% essentially they're doing five up to align:start position:0% essentially they're doing five up to five levels align:start position:0% five levels align:start position:0% five levels on each of their parameters and so align:start position:0% on each of their parameters and so align:start position:0% on each of their parameters and so they're using coded variables that we'll align:start position:0% they're using coded variables that we'll align:start position:0% they're using coded variables that we'll see in the later align:start position:0% see in the later align:start position:0% see in the later tables just like align:start position:0% tables just like align:start position:0% tables just like we're used to align:start position:0% we're used to align:start position:0% we're used to now align:start position:0% now align:start position:0% now of course the other benefit of doing a align:start position:0% of course the other benefit of doing a align:start position:0% of course the other benefit of doing a screening experiment align:start position:0% screening experiment align:start position:0% screening experiment is that also helps give you some align:start position:0% is that also helps give you some align:start position:0% is that also helps give you some information about reasonable design align:start position:0% information about reasonable design align:start position:0% information about reasonable design space for picking the range that you align:start position:0% space for picking the range that you align:start position:0% space for picking the range that you want to explore these parameters in align:start position:0% want to explore these parameters in align:start position:0% want to explore these parameters in otherwise they probably would have had align:start position:0% otherwise they probably would have had align:start position:0% otherwise they probably would have had to just go by by align:start position:0% to just go by by align:start position:0% to just go by by prior experience align:start position:0% prior experience align:start position:0% prior experience that's another advantage of doing sort align:start position:0% that's another advantage of doing sort align:start position:0% that's another advantage of doing sort of this align:start position:0% of this align:start position:0% of this two-stage experimentation with screening align:start position:0% two-stage experimentation with screening align:start position:0% two-stage experimentation with screening followed by a more extensive align:start position:0% followed by a more extensive align:start position:0% followed by a more extensive either a new doe or adding to the align:start position:0% either a new doe or adding to the align:start position:0% either a new doe or adding to the screening doe align:start position:0% screening doe align:start position:0% screening doe so they're going to be building second align:start position:0% so they're going to be building second align:start position:0% so they're going to be building second order polynomial models that is to say align:start position:0% order polynomial models that is to say align:start position:0% order polynomial models that is to say quadratic models with both align:start position:0% quadratic models with both align:start position:0% quadratic models with both interactions and then square terms align:start position:0% interactions and then square terms align:start position:0% interactions and then square terms just square terms align:start position:0% just square terms align:start position:0% just square terms in individual parameters align:start position:0% in individual parameters align:start position:0% in individual parameters they also mentioned that they're not align:start position:0% they also mentioned that they're not align:start position:0% they also mentioned that they're not going to do any transformations of the align:start position:0% going to do any transformations of the align:start position:0% going to do any transformations of the output align:start position:0% output align:start position:0% output variables such as a log or an inverse align:start position:0% align:start position:0% okay so we've already talked i think align:start position:0% okay so we've already talked i think align:start position:0% okay so we've already talked i think about those two those two points so in align:start position:0% about those two those two points so in align:start position:0% about those two those two points so in the in the paper they show the the align:start position:0% the in the paper they show the the align:start position:0% the in the paper they show the the variable uh test levels coded here's align:start position:0% variable uh test levels coded here's align:start position:0% variable uh test levels coded here's again our four inputs and the trials align:start position:0% again our four inputs and the trials align:start position:0% again our four inputs and the trials and align:start position:0% and align:start position:0% and uh align:start position:0% uh align:start position:0% uh if you look you can start to get a very align:start position:0% if you look you can start to get a very align:start position:0% if you look you can start to get a very nice sense of what this central align:start position:0% nice sense of what this central align:start position:0% nice sense of what this central composite looks like align:start position:0% composite looks like align:start position:0% composite looks like what they're essentially doing here align:start position:0% what they're essentially doing here align:start position:0% what they're essentially doing here breaking it up into let's call it three align:start position:0% breaking it up into let's call it three align:start position:0% breaking it up into let's call it three different segments align:start position:0% different segments align:start position:0% different segments of of design points align:start position:0% of of design points align:start position:0% of of design points what are these design points what do align:start position:0% what are these design points what do align:start position:0% what are these design points what do those correspond to align:start position:0% align:start position:0% there's a term we refer to those kind of align:start position:0% there's a term we refer to those kind of align:start position:0% there's a term we refer to those kind of design points as align:start position:0% align:start position:0% where does one of those design points align:start position:0% where does one of those design points align:start position:0% where does one of those design points lie align:start position:0% align:start position:0% take trial number one align:start position:0% align:start position:0% right it's way out minus two on the align:start position:0% right it's way out minus two on the align:start position:0% right it's way out minus two on the space axis align:start position:0% space axis align:start position:0% space axis and holding the others at their median align:start position:0% and holding the others at their median align:start position:0% and holding the others at their median or their their nominal value right align:start position:0% or their their nominal value right align:start position:0% or their their nominal value right so that's align:start position:0% so that's align:start position:0% so that's out on the end of one of the axes align:start position:0% out on the end of one of the axes align:start position:0% out on the end of one of the axes then the plus two minus two that's the align:start position:0% then the plus two minus two that's the align:start position:0% then the plus two minus two that's the other end of this so those are our axial align:start position:0% other end of this so those are our axial align:start position:0% other end of this so those are our axial points align:start position:0% points align:start position:0% points in a central composite design align:start position:0% align:start position:0% they're sort of the classic hold align:start position:0% they're sort of the classic hold align:start position:0% they're sort of the classic hold everything else constant vary one align:start position:0% everything else constant vary one align:start position:0% everything else constant vary one parameter at a time so they are giving align:start position:0% parameter at a time so they are giving align:start position:0% parameter at a time so they are giving you unambiguous sense of what that one align:start position:0% you unambiguous sense of what that one align:start position:0% you unambiguous sense of what that one parameter is doing but not able to align:start position:0% parameter is doing but not able to align:start position:0% parameter is doing but not able to capture interactions and so on they're align:start position:0% capture interactions and so on they're align:start position:0% capture interactions and so on they're also exploring sort of the outer limits align:start position:0% also exploring sort of the outer limits align:start position:0% also exploring sort of the outer limits of the design align:start position:0% of the design align:start position:0% of the design how about these points align:start position:0% align:start position:0% how many of them are there align:start position:0% align:start position:0% looks like there's 16 of them align:start position:0% align:start position:0% corner points exactly align:start position:0% corner points exactly align:start position:0% corner points exactly so those are all corner combinations of align:start position:0% so those are all corner combinations of align:start position:0% so those are all corner combinations of four parameters align:start position:0% four parameters align:start position:0% four parameters four squared 16 different corners align:start position:0% four squared 16 different corners align:start position:0% four squared 16 different corners okay align:start position:0% okay align:start position:0% okay and then finally these align:start position:0% and then finally these align:start position:0% and then finally these these are easy what are these align:start position:0% these are easy what are these align:start position:0% these are easy what are these center points right align:start position:0% center points right align:start position:0% center points right and so i don't know whether they have align:start position:0% and so i don't know whether they have align:start position:0% and so i don't know whether they have seven seven replicates align:start position:0% seven seven replicates align:start position:0% seven seven replicates so the other points are unreplicated align:start position:0% so the other points are unreplicated align:start position:0% so the other points are unreplicated by the way so you might be a little bit align:start position:0% by the way so you might be a little bit align:start position:0% by the way so you might be a little bit susceptible to to outliers in some of align:start position:0% susceptible to to outliers in some of align:start position:0% susceptible to to outliers in some of the other points but they have a fairly align:start position:0% the other points but they have a fairly align:start position:0% the other points but they have a fairly reasonable number seven replicates at align:start position:0% reasonable number seven replicates at align:start position:0% reasonable number seven replicates at the center so that'll allow some align:start position:0% align:start position:0% more or less reasonable estimate of some align:start position:0% more or less reasonable estimate of some align:start position:0% more or less reasonable estimate of some of the align:start position:0% of the align:start position:0% of the reproducibility some of the inherent align:start position:0% reproducibility some of the inherent align:start position:0% reproducibility some of the inherent variances in the process align:start position:0% align:start position:0% okay align:start position:0% okay align:start position:0% okay they say by the way that align:start position:0% they say by the way that align:start position:0% they say by the way that those replicates generally result in align:start position:0% those replicates generally result in align:start position:0% those replicates generally result in about one and a half to five percent one align:start position:0% about one and a half to five percent one align:start position:0% about one and a half to five percent one sigma align:start position:0% sigma align:start position:0% sigma range in the output align:start position:0% range in the output align:start position:0% range in the output parameters the paper says they did a align:start position:0% parameters the paper says they did a align:start position:0% parameters the paper says they did a randomized run order align:start position:0% randomized run order align:start position:0% randomized run order why would you do that align:start position:0% align:start position:0% why not just run you got trial one align:start position:0% why not just run you got trial one align:start position:0% why not just run you got trial one et cetera why not just run all the way align:start position:0% et cetera why not just run all the way align:start position:0% et cetera why not just run all the way down to align:start position:0% down to align:start position:0% down to run trial number 31 in this order align:start position:0% align:start position:0% to be independent align:start position:0% to be independent align:start position:0% to be independent uh align:start position:0% uh align:start position:0% uh independent of what align:start position:0% align:start position:0% pressure different appreciation align:start position:0% align:start position:0% uh align:start position:0% uh align:start position:0% uh i think you're right but but as you align:start position:0% i think you're right but but as you align:start position:0% i think you're right but but as you change these parameters that we're only align:start position:0% change these parameters that we're only align:start position:0% change these parameters that we're only playing with these these four parameters align:start position:0% playing with these these four parameters align:start position:0% playing with these these four parameters but there might still be noise other align:start position:0% but there might still be noise other align:start position:0% but there might still be noise other noise in the process or drifts like we align:start position:0% noise in the process or drifts like we align:start position:0% noise in the process or drifts like we talked about and it turns out in many of align:start position:0% talked about and it turns out in many of align:start position:0% talked about and it turns out in many of these deposition processes you actually align:start position:0% these deposition processes you actually align:start position:0% these deposition processes you actually have to worry about possible drifts align:start position:0% have to worry about possible drifts align:start position:0% have to worry about possible drifts in that you may get film build up on the align:start position:0% in that you may get film build up on the align:start position:0% in that you may get film build up on the inside of the reactor align:start position:0% inside of the reactor align:start position:0% inside of the reactor that over time might change a little bit align:start position:0% that over time might change a little bit align:start position:0% that over time might change a little bit the deposition process so by randomizing align:start position:0% the deposition process so by randomizing align:start position:0% the deposition process so by randomizing the run order you're randomizing the align:start position:0% the run order you're randomizing the align:start position:0% the run order you're randomizing the outputs and converting if there is a align:start position:0% outputs and converting if there is a align:start position:0% outputs and converting if there is a trend you basically convert that align:start position:0% trend you basically convert that align:start position:0% trend you basically convert that systematic trend into align:start position:0% systematic trend into align:start position:0% systematic trend into noise align:start position:0% noise align:start position:0% noise it may still perturb your values but now align:start position:0% it may still perturb your values but now align:start position:0% it may still perturb your values but now it does it kind of distributed in a align:start position:0% it does it kind of distributed in a align:start position:0% it does it kind of distributed in a random fashion across all of your align:start position:0% random fashion across all of your align:start position:0% random fashion across all of your outputs and it will add to the noise the align:start position:0% outputs and it will add to the noise the align:start position:0% outputs and it will add to the noise the change the reproducibility values align:start position:0% change the reproducibility values align:start position:0% change the reproducibility values but at least then it's not align:start position:0% but at least then it's not align:start position:0% but at least then it's not fooling you in a systematic way about align:start position:0% fooling you in a systematic way about align:start position:0% fooling you in a systematic way about confusing if there is that kind of other align:start position:0% confusing if there is that kind of other align:start position:0% confusing if there is that kind of other drift it will not confuse you or alias align:start position:0% drift it will not confuse you or alias align:start position:0% drift it will not confuse you or alias with some other parameter align:start position:0% with some other parameter align:start position:0% with some other parameter okay align:start position:0% okay align:start position:0% okay i have one minor annoyance in the paper align:start position:0% i have one minor annoyance in the paper align:start position:0% i have one minor annoyance in the paper and so if you're writing papers and align:start position:0% and so if you're writing papers and align:start position:0% and so if you're writing papers and you're you do a randomized run align:start position:0% you're you do a randomized run align:start position:0% you're you do a randomized run uh align:start position:0% uh align:start position:0% uh tell me what the run order actually was align:start position:0% tell me what the run order actually was align:start position:0% tell me what the run order actually was if you're telling me the data i would align:start position:0% if you're telling me the data i would align:start position:0% if you're telling me the data i would love to actually know the run order align:start position:0% love to actually know the run order align:start position:0% love to actually know the run order because then when i do residual analysis align:start position:0% because then when i do residual analysis align:start position:0% because then when i do residual analysis and look at the residuals as a function align:start position:0% and look at the residuals as a function align:start position:0% and look at the residuals as a function of time align:start position:0% of time align:start position:0% of time i can actually align:start position:0% i can actually align:start position:0% i can actually go in and see if there was perhaps a align:start position:0% go in and see if there was perhaps a align:start position:0% go in and see if there was perhaps a systematic time drift align:start position:0% systematic time drift align:start position:0% systematic time drift and it really doesn't cost much align:start position:0% and it really doesn't cost much align:start position:0% and it really doesn't cost much to do that in align:start position:0% to do that in align:start position:0% to do that in ink space in the paper align:start position:0% ink space in the paper align:start position:0% ink space in the paper so for example here they had trial some align:start position:0% so for example here they had trial some align:start position:0% so for example here they had trial some papers you will actually see they go in align:start position:0% papers you will actually see they go in align:start position:0% papers you will actually see they go in and might put align:start position:0% and might put align:start position:0% and might put a number sort of right beside it align:start position:0% a number sort of right beside it align:start position:0% a number sort of right beside it in parentheses or something like that or align:start position:0% in parentheses or something like that or align:start position:0% in parentheses or something like that or actual another align:start position:0% actual another align:start position:0% actual another column in the table or whatever that align:start position:0% column in the table or whatever that align:start position:0% column in the table or whatever that tells me run order not just trial number align:start position:0% tells me run order not just trial number align:start position:0% tells me run order not just trial number okay so do that align:start position:0% okay so do that align:start position:0% okay so do that if you write write papers with data in align:start position:0% if you write write papers with data in align:start position:0% if you write write papers with data in it align:start position:0% it align:start position:0% it they also say they do did some outlier align:start position:0% they also say they do did some outlier align:start position:0% they also say they do did some outlier analysis they don't really uh align:start position:0% analysis they don't really uh align:start position:0% analysis they don't really uh say what that was but they know they do align:start position:0% say what that was but they know they do align:start position:0% say what that was but they know they do note it in the data align:start position:0% note it in the data align:start position:0% note it in the data and by the way align:start position:0% and by the way align:start position:0% and by the way oh i may not have put this on the align:start position:0% oh i may not have put this on the align:start position:0% oh i may not have put this on the site yet i will also add the spreadsheet align:start position:0% site yet i will also add the spreadsheet align:start position:0% site yet i will also add the spreadsheet with this basic data so if you want to align:start position:0% with this basic data so if you want to align:start position:0% with this basic data so if you want to grab this later and play with the data align:start position:0% grab this later and play with the data align:start position:0% grab this later and play with the data and do some of your own response surface align:start position:0% and do some of your own response surface align:start position:0% and do some of your own response surface model fits with align:start position:0% model fits with align:start position:0% model fits with jump or some of the other tools you can align:start position:0% jump or some of the other tools you can align:start position:0% jump or some of the other tools you can do that align:start position:0% do that align:start position:0% do that here's the output align:start position:0% here's the output align:start position:0% here's the output trial number and then each of the seven align:start position:0% trial number and then each of the seven align:start position:0% trial number and then each of the seven different outputs so these are the seven align:start position:0% different outputs so these are the seven align:start position:0% different outputs so these are the seven different outputs and basically just the align:start position:0% different outputs and basically just the align:start position:0% different outputs and basically just the raw the raw measured data align:start position:0% raw the raw measured data align:start position:0% raw the raw measured data they align:start position:0% they align:start position:0% they also may have asterisked or footnoted a align:start position:0% also may have asterisked or footnoted a align:start position:0% also may have asterisked or footnoted a few of these which is a good practice align:start position:0% few of these which is a good practice align:start position:0% few of these which is a good practice they said you know align:start position:0% they said you know align:start position:0% they said you know weird things like align:start position:0% align:start position:0% indicates entries determined to be align:start position:0% indicates entries determined to be align:start position:0% indicates entries determined to be outliers align:start position:0% outliers align:start position:0% outliers or if they knew something weird happened align:start position:0% or if they knew something weird happened align:start position:0% or if they knew something weird happened with the measurement of that particular align:start position:0% with the measurement of that particular align:start position:0% with the measurement of that particular data point or that entire row align:start position:0% data point or that entire row align:start position:0% data point or that entire row they would note that that can be really align:start position:0% they would note that that can be really align:start position:0% they would note that that can be really critical when you're interpreting the align:start position:0% critical when you're interpreting the align:start position:0% critical when you're interpreting the data align:start position:0% data align:start position:0% data then what they've done down here at the align:start position:0% then what they've done down here at the align:start position:0% then what they've done down here at the bottom is simply calculated align:start position:0% bottom is simply calculated align:start position:0% bottom is simply calculated things like align:start position:0% things like align:start position:0% things like the standard deviation align:start position:0% the standard deviation align:start position:0% the standard deviation across align:start position:0% across align:start position:0% across across all of those align:start position:0% across all of those align:start position:0% across all of those i think that's across align:start position:0% i think that's across align:start position:0% i think that's across just the replicate runs you can see that align:start position:0% just the replicate runs you can see that align:start position:0% just the replicate runs you can see that underneath here align:start position:0% underneath here align:start position:0% underneath here yeah okay align:start position:0% yeah okay align:start position:0% yeah okay so it gives you a sense of the align:start position:0% so it gives you a sense of the align:start position:0% so it gives you a sense of the percentage variance one sigma percentage align:start position:0% percentage variance one sigma percentage align:start position:0% percentage variance one sigma percentage variance in each of the output align:start position:0% variance in each of the output align:start position:0% variance in each of the output parameters align:start position:0% parameters align:start position:0% parameters yes question align:start position:0% yes question align:start position:0% yes question if one of the entries was align:start position:0% if one of the entries was align:start position:0% if one of the entries was as an issue doesn't that mean align:start position:0% as an issue doesn't that mean align:start position:0% as an issue doesn't that mean the whole row align:start position:0% the whole row align:start position:0% the whole row as an issue align:start position:0% as an issue align:start position:0% as an issue so shouldn't that whole row be discarded align:start position:0% so shouldn't that whole row be discarded align:start position:0% so shouldn't that whole row be discarded how can you for example you know in run align:start position:0% how can you for example you know in run align:start position:0% how can you for example you know in run six have an issue with let's say 8.1 and align:start position:0% six have an issue with let's say 8.1 and align:start position:0% six have an issue with let's say 8.1 and say you know this looks weird align:start position:0% say you know this looks weird align:start position:0% say you know this looks weird and you're not going to align:start position:0% and you're not going to align:start position:0% and you're not going to um use it in the row uh align:start position:0% um use it in the row uh align:start position:0% um use it in the row uh so align:start position:0% align:start position:0% how can you have just an issue with one align:start position:0% how can you have just an issue with one align:start position:0% how can you have just an issue with one element and not an issue with the whole align:start position:0% element and not an issue with the whole align:start position:0% element and not an issue with the whole row align:start position:0% row align:start position:0% row basically i think it comes down to some align:start position:0% basically i think it comes down to some align:start position:0% basically i think it comes down to some engineering knowledge on what the issue align:start position:0% engineering knowledge on what the issue align:start position:0% engineering knowledge on what the issue may actually be so in particular if align:start position:0% may actually be so in particular if align:start position:0% may actually be so in particular if you're worried that align:start position:0% you're worried that align:start position:0% you're worried that there's something wrong with the align:start position:0% there's something wrong with the align:start position:0% there's something wrong with the measurement align:start position:0% measurement align:start position:0% measurement then that does not necessarily align:start position:0% then that does not necessarily align:start position:0% then that does not necessarily contaminate your measurements of other align:start position:0% contaminate your measurements of other align:start position:0% contaminate your measurements of other other align:start position:0% other align:start position:0% other other columns align:start position:0% other columns align:start position:0% other columns if you think it's really an indicator align:start position:0% if you think it's really an indicator align:start position:0% if you think it's really an indicator that the whole run align:start position:0% that the whole run align:start position:0% that the whole run or the whole trial was align:start position:0% or the whole trial was align:start position:0% or the whole trial was was messed up somehow align:start position:0% was messed up somehow align:start position:0% was messed up somehow then i agree you would probably want to align:start position:0% then i agree you would probably want to align:start position:0% then i agree you would probably want to to omit that entire run from your whole align:start position:0% to omit that entire run from your whole align:start position:0% to omit that entire run from your whole table or maybe repeat that run align:start position:0% table or maybe repeat that run align:start position:0% table or maybe repeat that run in this case with single wafer align:start position:0% in this case with single wafer align:start position:0% in this case with single wafer processing is probably an easy thing to align:start position:0% processing is probably an easy thing to align:start position:0% processing is probably an easy thing to do align:start position:0% do align:start position:0% do i guess the second question if you're align:start position:0% i guess the second question if you're align:start position:0% i guess the second question if you're doing experiments align:start position:0% doing experiments align:start position:0% doing experiments that's an issue align:start position:0% that's an issue align:start position:0% that's an issue one of the numbers looks align:start position:0% one of the numbers looks align:start position:0% one of the numbers looks funky yeah um align:start position:0% funky yeah um align:start position:0% funky yeah um the fact and we want to eliminate that align:start position:0% the fact and we want to eliminate that align:start position:0% the fact and we want to eliminate that whole row what does that mean for align:start position:0% whole row what does that mean for align:start position:0% whole row what does that mean for our experiment if it's for example like align:start position:0% our experiment if it's for example like align:start position:0% our experiment if it's for example like the only we didn't do replicas is the align:start position:0% the only we didn't do replicas is the align:start position:0% the only we didn't do replicas is the center point and now we don't have a align:start position:0% center point and now we don't have a align:start position:0% center point and now we don't have a center point yeah so the other question align:start position:0% center point yeah so the other question align:start position:0% center point yeah so the other question is if you do have a bad data row how align:start position:0% is if you do have a bad data row how align:start position:0% is if you do have a bad data row how does that mess up your data analysis align:start position:0% does that mess up your data analysis align:start position:0% does that mess up your data analysis and and align:start position:0% and and align:start position:0% and and the answer is depending on the kind of align:start position:0% the answer is depending on the kind of align:start position:0% the answer is depending on the kind of doe that you have and the kinds of align:start position:0% doe that you have and the kinds of align:start position:0% doe that you have and the kinds of analysis that you apply align:start position:0% analysis that you apply align:start position:0% analysis that you apply you can be more or less robust to those align:start position:0% you can be more or less robust to those align:start position:0% you can be more or less robust to those kinds of missing data points align:start position:0% kinds of missing data points align:start position:0% kinds of missing data points one of the nice things about a central align:start position:0% one of the nice things about a central align:start position:0% one of the nice things about a central composite design align:start position:0% composite design align:start position:0% composite design and especially one where they've got you align:start position:0% and especially one where they've got you align:start position:0% and especially one where they've got you know kind of this these five levels is align:start position:0% know kind of this these five levels is align:start position:0% know kind of this these five levels is if you do that in conjunction with align:start position:0% if you do that in conjunction with align:start position:0% if you do that in conjunction with response surface modeling regression align:start position:0% response surface modeling regression align:start position:0% response surface modeling regression modeling in particular the numerical align:start position:0% modeling in particular the numerical align:start position:0% modeling in particular the numerical regression modeling you're fairly well align:start position:0% regression modeling you're fairly well align:start position:0% regression modeling you're fairly well off if you have a few missing data align:start position:0% off if you have a few missing data align:start position:0% off if you have a few missing data points it would throw you for a loop in align:start position:0% points it would throw you for a loop in align:start position:0% points it would throw you for a loop in some of the shorthand formation of align:start position:0% some of the shorthand formation of align:start position:0% some of the shorthand formation of contrast that we talked about because align:start position:0% contrast that we talked about because align:start position:0% contrast that we talked about because then you know your typical averaging align:start position:0% then you know your typical averaging align:start position:0% then you know your typical averaging that's implicit and some of those are align:start position:0% that's implicit and some of those are align:start position:0% that's implicit and some of those are wrong align:start position:0% wrong align:start position:0% wrong so you've got to be able but but if you align:start position:0% so you've got to be able but but if you align:start position:0% so you've got to be able but but if you do align:start position:0% do align:start position:0% do regression modeling in conjunction with align:start position:0% regression modeling in conjunction with align:start position:0% regression modeling in conjunction with a fairly rich align:start position:0% align:start position:0% sort of align:start position:0% sort of align:start position:0% sort of extra data points are lurking in there align:start position:0% extra data points are lurking in there align:start position:0% extra data points are lurking in there extra degrees of freedom are lurking in align:start position:0% extra degrees of freedom are lurking in align:start position:0% extra degrees of freedom are lurking in there that still allow you to reliably align:start position:0% there that still allow you to reliably align:start position:0% there that still allow you to reliably fit models align:start position:0% align:start position:0% so the paper says the anova is performed align:start position:0% so the paper says the anova is performed align:start position:0% so the paper says the anova is performed but it is not shown align:start position:0% but it is not shown align:start position:0% but it is not shown i want to see the anova table so well we align:start position:0% i want to see the anova table so well we align:start position:0% i want to see the anova table so well we can basically generate those on our own align:start position:0% can basically generate those on our own align:start position:0% can basically generate those on our own and so in this case align:start position:0% and so in this case align:start position:0% and so in this case if the paper has the raw data in it i align:start position:0% if the paper has the raw data in it i align:start position:0% if the paper has the raw data in it i think it's okay that they didn't show align:start position:0% think it's okay that they didn't show align:start position:0% think it's okay that they didn't show the anova because i can regenerate that align:start position:0% the anova because i can regenerate that align:start position:0% the anova because i can regenerate that papers i hate are they don't show you align:start position:0% papers i hate are they don't show you align:start position:0% papers i hate are they don't show you the data and they don't show you the align:start position:0% the data and they don't show you the align:start position:0% the data and they don't show you the anova and so you don't know really what align:start position:0% anova and so you don't know really what align:start position:0% anova and so you don't know really what the noise was you don't know align:start position:0% the noise was you don't know align:start position:0% the noise was you don't know significances of parameters you don't align:start position:0% significances of parameters you don't align:start position:0% significances of parameters you don't know which parameters often they won't align:start position:0% know which parameters often they won't align:start position:0% know which parameters often they won't tell you really the significance level align:start position:0% tell you really the significance level align:start position:0% tell you really the significance level that was used to accept a model all of align:start position:0% that was used to accept a model all of align:start position:0% that was used to accept a model all of those sorts of things so in this case align:start position:0% those sorts of things so in this case align:start position:0% those sorts of things so in this case i'm i'm happy with the paper because align:start position:0% i'm i'm happy with the paper because align:start position:0% i'm i'm happy with the paper because i've got the raw data align:start position:0% i've got the raw data align:start position:0% i've got the raw data they claim that each of the model is align:start position:0% they claim that each of the model is align:start position:0% they claim that each of the model is significant at 99.9 percent uh align:start position:0% significant at 99.9 percent uh align:start position:0% significant at 99.9 percent uh significance level that's just align:start position:0% significance level that's just align:start position:0% significance level that's just their their p-value on the f-test for align:start position:0% their their p-value on the f-test for align:start position:0% their their p-value on the f-test for the whole model align:start position:0% the whole model align:start position:0% the whole model so these are very align:start position:0% so these are very align:start position:0% so these are very uh at least there is some parameter that align:start position:0% uh at least there is some parameter that align:start position:0% uh at least there is some parameter that is significant in influencing the output align:start position:0% is significant in influencing the output align:start position:0% is significant in influencing the output they report also r squareds which are is align:start position:0% they report also r squareds which are is align:start position:0% they report also r squareds which are is nice and for most of the models they are align:start position:0% nice and for most of the models they are align:start position:0% nice and for most of the models they are really quite high in the 0.88 to 0.97 align:start position:0% really quite high in the 0.88 to 0.97 align:start position:0% really quite high in the 0.88 to 0.97 range align:start position:0% range align:start position:0% range the lowest one was align:start position:0% the lowest one was align:start position:0% the lowest one was resistivity at about 0.79 align:start position:0% resistivity at about 0.79 align:start position:0% resistivity at about 0.79 okay align:start position:0% align:start position:0% they also report some lack of fit align:start position:0% they also report some lack of fit align:start position:0% they also report some lack of fit mention a little bit of evidence of lack align:start position:0% mention a little bit of evidence of lack align:start position:0% mention a little bit of evidence of lack of fit for uniformity and reflectance align:start position:0% of fit for uniformity and reflectance align:start position:0% of fit for uniformity and reflectance this align:start position:0% this align:start position:0% this lack of fit for sheet resistance align:start position:0% lack of fit for sheet resistance align:start position:0% lack of fit for sheet resistance uniformity align:start position:0% uniformity align:start position:0% uniformity is really interesting align:start position:0% is really interesting align:start position:0% is really interesting and in fact so interesting align:start position:0% and in fact so interesting align:start position:0% and in fact so interesting that i think a week from today if dave align:start position:0% that i think a week from today if dave align:start position:0% that i think a week from today if dave is talking on thursday align:start position:0% is talking on thursday align:start position:0% is talking on thursday i will spend a whole lecture align:start position:0% i will spend a whole lecture align:start position:0% i will spend a whole lecture on modeling of uniformity align:start position:0% on modeling of uniformity align:start position:0% on modeling of uniformity response surface modeling approaches align:start position:0% response surface modeling approaches align:start position:0% response surface modeling approaches spatial modeling approaches appropriate align:start position:0% spatial modeling approaches appropriate align:start position:0% spatial modeling approaches appropriate for uniformity align:start position:0% for uniformity align:start position:0% for uniformity the preview is align:start position:0% the preview is align:start position:0% the preview is think of the calculations that go into a align:start position:0% think of the calculations that go into a align:start position:0% think of the calculations that go into a wafer level uniformity calcu align:start position:0% wafer level uniformity calcu align:start position:0% wafer level uniformity calcu you know the calculation of that align:start position:0% you know the calculation of that align:start position:0% you know the calculation of that you might measure i think they're align:start position:0% you might measure i think they're align:start position:0% you might measure i think they're measuring 49 points across the wafer align:start position:0% measuring 49 points across the wafer align:start position:0% measuring 49 points across the wafer they're do it forming sums of squared align:start position:0% they're do it forming sums of squared align:start position:0% they're do it forming sums of squared deviations and then taking a square root align:start position:0% deviations and then taking a square root align:start position:0% deviations and then taking a square root and then perhaps normalizing by some align:start position:0% and then perhaps normalizing by some align:start position:0% and then perhaps normalizing by some mean align:start position:0% mean align:start position:0% mean does that sound like a set of linear align:start position:0% does that sound like a set of linear align:start position:0% does that sound like a set of linear operations align:start position:0% operations align:start position:0% operations no that there's a lot of implicit align:start position:0% no that there's a lot of implicit align:start position:0% no that there's a lot of implicit non-linearity align:start position:0% non-linearity align:start position:0% non-linearity lurking just in the calculation of that align:start position:0% lurking just in the calculation of that align:start position:0% lurking just in the calculation of that of that metric align:start position:0% of that metric align:start position:0% of that metric and so it's not surprising that there's align:start position:0% and so it's not surprising that there's align:start position:0% and so it's not surprising that there's quadratic lack of fit or inability of a align:start position:0% quadratic lack of fit or inability of a align:start position:0% quadratic lack of fit or inability of a quadratic model to deal with a align:start position:0% quadratic model to deal with a align:start position:0% quadratic model to deal with a uniformity metric align:start position:0% uniformity metric align:start position:0% uniformity metric a spatial uniformity metric and we'll align:start position:0% a spatial uniformity metric and we'll align:start position:0% a spatial uniformity metric and we'll talk more about that next week but align:start position:0% talk more about that next week but align:start position:0% talk more about that next week but but um align:start position:0% but um align:start position:0% but um so they they do mention this uh at least align:start position:0% so they they do mention this uh at least align:start position:0% so they they do mention this uh at least they also mentioned some some potential align:start position:0% they also mentioned some some potential align:start position:0% they also mentioned some some potential lack of fit and reflectance i don't have align:start position:0% lack of fit and reflectance i don't have align:start position:0% lack of fit and reflectance i don't have an explanation for that they conjecture align:start position:0% an explanation for that they conjecture align:start position:0% an explanation for that they conjecture it's due to a very small pure error term align:start position:0% it's due to a very small pure error term align:start position:0% it's due to a very small pure error term so that you're able to detect align:start position:0% so that you're able to detect align:start position:0% so that you're able to detect with pretty fine resolution if there's align:start position:0% with pretty fine resolution if there's align:start position:0% with pretty fine resolution if there's any subtle non-quadratic align:start position:0% any subtle non-quadratic align:start position:0% any subtle non-quadratic uh trends but they're not concerned with align:start position:0% uh trends but they're not concerned with align:start position:0% uh trends but they're not concerned with it it's align:start position:0% it it's align:start position:0% it it's they still feel that they have very good align:start position:0% they still feel that they have very good align:start position:0% they still feel that they have very good good models align:start position:0% good models align:start position:0% good models it's also the case align:start position:0% it's also the case align:start position:0% it's also the case that there may be transformations align:start position:0% that there may be transformations align:start position:0% that there may be transformations that could be applied to improve these align:start position:0% that could be applied to improve these align:start position:0% that could be applied to improve these models align:start position:0% models align:start position:0% models and that also for the sheet resistivity align:start position:0% and that also for the sheet resistivity align:start position:0% and that also for the sheet resistivity or sheet resistance measurement align:start position:0% or sheet resistance measurement align:start position:0% or sheet resistance measurement there might be natural transformations align:start position:0% there might be natural transformations align:start position:0% there might be natural transformations that you might think that apply align:start position:0% that you might think that apply align:start position:0% that you might think that apply in fact if this is related to things align:start position:0% in fact if this is related to things align:start position:0% in fact if this is related to things like align:start position:0% like align:start position:0% like resistance divided by a thickness align:start position:0% resistance divided by a thickness align:start position:0% resistance divided by a thickness and the thickness align:start position:0% and the thickness align:start position:0% and the thickness grows linearly align:start position:0% grows linearly align:start position:0% grows linearly with align:start position:0% with align:start position:0% with you know maybe you have a linear align:start position:0% you know maybe you have a linear align:start position:0% you know maybe you have a linear dependence on on align:start position:0% dependence on on align:start position:0% dependence on on uh growth rate or something like that align:start position:0% uh growth rate or something like that align:start position:0% uh growth rate or something like that there may be inherent in there some one align:start position:0% there may be inherent in there some one align:start position:0% there may be inherent in there some one over kinds of transformations that might align:start position:0% over kinds of transformations that might align:start position:0% over kinds of transformations that might actually be a better thing to do to take align:start position:0% actually be a better thing to do to take align:start position:0% actually be a better thing to do to take that sheet resistance invert it and align:start position:0% that sheet resistance invert it and align:start position:0% that sheet resistance invert it and model the inverse of sheet resistance as align:start position:0% model the inverse of sheet resistance as align:start position:0% model the inverse of sheet resistance as a function of the input parameters align:start position:0% a function of the input parameters align:start position:0% a function of the input parameters so you can play around with that and align:start position:0% so you can play around with that and align:start position:0% so you can play around with that and actually that's a neat fun thing to play align:start position:0% actually that's a neat fun thing to play align:start position:0% actually that's a neat fun thing to play around with with some of these models align:start position:0% around with with some of these models align:start position:0% around with with some of these models they didn't do that align:start position:0% they didn't do that align:start position:0% they didn't do that but you could align:start position:0% but you could align:start position:0% but you could and then they do show the regression align:start position:0% and then they do show the regression align:start position:0% and then they do show the regression models here's the output these are the align:start position:0% models here's the output these are the align:start position:0% models here's the output these are the regression coefficients all of the align:start position:0% regression coefficients all of the align:start position:0% regression coefficients all of the linear terms and the cross terms and the align:start position:0% linear terms and the cross terms and the align:start position:0% linear terms and the cross terms and the square terms and what they've done in align:start position:0% square terms and what they've done in align:start position:0% square terms and what they've done in this table is report the align:start position:0% this table is report the align:start position:0% this table is report the the fit for just the significant terms align:start position:0% the fit for just the significant terms align:start position:0% the fit for just the significant terms if the coefficient was align:start position:0% if the coefficient was align:start position:0% if the coefficient was determined in their stepwise regression align:start position:0% determined in their stepwise regression align:start position:0% determined in their stepwise regression to be insignificant they've left it off align:start position:0% to be insignificant they've left it off align:start position:0% to be insignificant they've left it off i don't think they actually told us align:start position:0% i don't think they actually told us align:start position:0% i don't think they actually told us what their criteria for inclusion what align:start position:0% what their criteria for inclusion what align:start position:0% what their criteria for inclusion what their significance level cut off align:start position:0% their significance level cut off align:start position:0% their significance level cut off was align:start position:0% was align:start position:0% was in here so that would be a good thing to align:start position:0% in here so that would be a good thing to align:start position:0% in here so that would be a good thing to do as well but you can see lots of lots align:start position:0% do as well but you can see lots of lots align:start position:0% do as well but you can see lots of lots of models lots of align:start position:0% of models lots of align:start position:0% of models lots of uh second order terms these are you know align:start position:0% uh second order terms these are you know align:start position:0% uh second order terms these are you know interactions align:start position:0% interactions align:start position:0% interactions b 1 1 b 2 2 b 3 3 b align:start position:0% b 1 1 b 2 2 b 3 3 b align:start position:0% b 1 1 b 2 2 b 3 3 b those are pure square terms and then align:start position:0% those are pure square terms and then align:start position:0% those are pure square terms and then these are interactions so there's a fair align:start position:0% these are interactions so there's a fair align:start position:0% these are interactions so there's a fair amount of square activity going on align:start position:0% amount of square activity going on align:start position:0% amount of square activity going on as well as interactions align:start position:0% as well as interactions align:start position:0% as well as interactions they also then use those fit models to align:start position:0% they also then use those fit models to align:start position:0% they also then use those fit models to start to form align:start position:0% start to form align:start position:0% start to form contour plots and this is how they start align:start position:0% contour plots and this is how they start align:start position:0% contour plots and this is how they start to make then some engineering judgment align:start position:0% to make then some engineering judgment align:start position:0% to make then some engineering judgment about different trends so they would align:start position:0% about different trends so they would align:start position:0% about different trends so they would plot align:start position:0% align:start position:0% realize they've got four inputs and align:start position:0% realize they've got four inputs and align:start position:0% realize they've got four inputs and seven outputs you possibly can't align:start position:0% seven outputs you possibly can't align:start position:0% seven outputs you possibly can't possibly have a align:start position:0% possibly have a align:start position:0% possibly have a whatever you know four-dimensional input align:start position:0% whatever you know four-dimensional input align:start position:0% whatever you know four-dimensional input space is very hard to visualize so what align:start position:0% space is very hard to visualize so what align:start position:0% space is very hard to visualize so what you resort to are typically either two align:start position:0% you resort to are typically either two align:start position:0% you resort to are typically either two typically sort of two dimensional plots align:start position:0% typically sort of two dimensional plots align:start position:0% typically sort of two dimensional plots where you might plot one output or two align:start position:0% where you might plot one output or two align:start position:0% where you might plot one output or two outputs as a contour so in this align:start position:0% outputs as a contour so in this align:start position:0% outputs as a contour so in this particular plot the solid line i believe align:start position:0% particular plot the solid line i believe align:start position:0% particular plot the solid line i believe is growth rate align:start position:0% is growth rate align:start position:0% is growth rate and the dashed line here is wf6 align:start position:0% and the dashed line here is wf6 align:start position:0% and the dashed line here is wf6 conversion align:start position:0% align:start position:0% and you can start to feel get a feel for align:start position:0% and you can start to feel get a feel for align:start position:0% and you can start to feel get a feel for as for example align:start position:0% as for example align:start position:0% as for example the wf-6 pressure and the temperature align:start position:0% the wf-6 pressure and the temperature align:start position:0% the wf-6 pressure and the temperature change align:start position:0% change align:start position:0% change what trends you get in those two align:start position:0% what trends you get in those two align:start position:0% what trends you get in those two parameters and what trade-offs you get align:start position:0% parameters and what trade-offs you get align:start position:0% parameters and what trade-offs you get in those two parameters so things like align:start position:0% in those two parameters so things like align:start position:0% in those two parameters so things like as the temperature increases align:start position:0% as the temperature increases align:start position:0% as the temperature increases so temperature increasing this way the align:start position:0% so temperature increasing this way the align:start position:0% so temperature increasing this way the growth rate align:start position:0% growth rate align:start position:0% growth rate kind of increases align:start position:0% kind of increases align:start position:0% kind of increases right so the growth rate is the solid align:start position:0% right so the growth rate is the solid align:start position:0% right so the growth rate is the solid line align:start position:0% line align:start position:0% line so pretty directly align:start position:0% so pretty directly align:start position:0% so pretty directly with temperature that looks like a align:start position:0% with temperature that looks like a align:start position:0% with temperature that looks like a fairly strong direct impact as you might align:start position:0% fairly strong direct impact as you might align:start position:0% fairly strong direct impact as you might expect with a chemically align:start position:0% expect with a chemically align:start position:0% expect with a chemically dependent kind of process align:start position:0% dependent kind of process align:start position:0% dependent kind of process so they can start to get a feel for what align:start position:0% so they can start to get a feel for what align:start position:0% so they can start to get a feel for what some of these trends are and look at align:start position:0% some of these trends are and look at align:start position:0% some of these trends are and look at what align:start position:0% what align:start position:0% what for example has the biggest impact on align:start position:0% for example has the biggest impact on align:start position:0% for example has the biggest impact on growth rate of these of these parameters align:start position:0% growth rate of these of these parameters align:start position:0% growth rate of these of these parameters and which have align:start position:0% and which have align:start position:0% and which have either insignificant or small in align:start position:0% either insignificant or small in align:start position:0% either insignificant or small in magnitude align:start position:0% magnitude align:start position:0% magnitude kinds of effects align:start position:0% kinds of effects align:start position:0% kinds of effects now they can also start to do things align:start position:0% now they can also start to do things align:start position:0% now they can also start to do things like overlay requirements or conditions align:start position:0% like overlay requirements or conditions align:start position:0% like overlay requirements or conditions on these to start to understand the align:start position:0% on these to start to understand the align:start position:0% on these to start to understand the feasible operating space for process align:start position:0% feasible operating space for process align:start position:0% feasible operating space for process design and optimization so for example align:start position:0% design and optimization so for example align:start position:0% design and optimization so for example if your growth rate had to be 500 align:start position:0% if your growth rate had to be 500 align:start position:0% if your growth rate had to be 500 nanometers per minute or higher align:start position:0% nanometers per minute or higher align:start position:0% nanometers per minute or higher you might then have whoops align:start position:0% you might then have whoops align:start position:0% you might then have whoops you know kind of shading on this plot align:start position:0% you know kind of shading on this plot align:start position:0% you know kind of shading on this plot you would say okay i really need to be align:start position:0% you would say okay i really need to be align:start position:0% you would say okay i really need to be above that 500 nanometer per minute align:start position:0% above that 500 nanometer per minute align:start position:0% above that 500 nanometer per minute contour and that tells me align:start position:0% contour and that tells me align:start position:0% contour and that tells me where i need to be at least with respect align:start position:0% where i need to be at least with respect align:start position:0% where i need to be at least with respect to those two parameters in terms of align:start position:0% to those two parameters in terms of align:start position:0% to those two parameters in terms of feasible operating space align:start position:0% feasible operating space align:start position:0% feasible operating space so already it tells you a lot just align:start position:0% so already it tells you a lot just align:start position:0% so already it tells you a lot just looking at these plots of where you align:start position:0% looking at these plots of where you align:start position:0% looking at these plots of where you might need to be in order to to achieve align:start position:0% might need to be in order to to achieve align:start position:0% might need to be in order to to achieve your multiple goals it may not give you align:start position:0% your multiple goals it may not give you align:start position:0% your multiple goals it may not give you a numerical optimum point align:start position:0% a numerical optimum point align:start position:0% a numerical optimum point but align:start position:0% but align:start position:0% but all by itself lots of looking at the align:start position:0% all by itself lots of looking at the align:start position:0% all by itself lots of looking at the data and plotting of the data is really align:start position:0% data and plotting of the data is really align:start position:0% data and plotting of the data is really valuable align:start position:0% valuable align:start position:0% valuable they do that for not only growth rate align:start position:0% they do that for not only growth rate align:start position:0% they do that for not only growth rate and wf6 but things like resistivity align:start position:0% and wf6 but things like resistivity align:start position:0% and wf6 but things like resistivity reflectance align:start position:0% reflectance align:start position:0% reflectance they start to look and see you know is align:start position:0% they start to look and see you know is align:start position:0% they start to look and see you know is the overall design going to be feasible align:start position:0% the overall design going to be feasible align:start position:0% the overall design going to be feasible what has the biggest effect on align:start position:0% what has the biggest effect on align:start position:0% what has the biggest effect on morphology align:start position:0% morphology align:start position:0% morphology those sorts of things align:start position:0% align:start position:0% we found uh align:start position:0% we found uh align:start position:0% we found uh if you look at their contour plots most align:start position:0% if you look at their contour plots most align:start position:0% if you look at their contour plots most of these contour plots are are align:start position:0% of these contour plots are are align:start position:0% of these contour plots are are reasonably align:start position:0% reasonably align:start position:0% reasonably you know these are nice align:start position:0% you know these are nice align:start position:0% you know these are nice semi-linear kinds of plots align:start position:0% semi-linear kinds of plots align:start position:0% semi-linear kinds of plots their uniformity plot these look bizarre align:start position:0% their uniformity plot these look bizarre align:start position:0% their uniformity plot these look bizarre right very complex surfaces align:start position:0% right very complex surfaces align:start position:0% right very complex surfaces and again that's a reflection that this align:start position:0% and again that's a reflection that this align:start position:0% and again that's a reflection that this uniformity metric the spatial uniform align:start position:0% uniformity metric the spatial uniform align:start position:0% uniformity metric the spatial uniform internet is a very complex function of align:start position:0% internet is a very complex function of align:start position:0% internet is a very complex function of the input parameters align:start position:0% align:start position:0% what they then do finally to round out align:start position:0% what they then do finally to round out align:start position:0% what they then do finally to round out the paper is talk a little bit about align:start position:0% the paper is talk a little bit about align:start position:0% the paper is talk a little bit about optimization so what they will have is align:start position:0% optimization so what they will have is align:start position:0% optimization so what they will have is some of those limits like i i said for align:start position:0% some of those limits like i i said for align:start position:0% some of those limits like i i said for example on the growth rate i guess i align:start position:0% example on the growth rate i guess i align:start position:0% example on the growth rate i guess i guess it wasn't a 500 nanometer cut off align:start position:0% guess it wasn't a 500 nanometer cut off align:start position:0% guess it wasn't a 500 nanometer cut off they needed at least 300 nanometer but align:start position:0% they needed at least 300 nanometer but align:start position:0% they needed at least 300 nanometer but they would like to get that align:start position:0% they would like to get that align:start position:0% they would like to get that higher is better and they would align:start position:0% higher is better and they would align:start position:0% higher is better and they would basically have different align:start position:0% basically have different align:start position:0% basically have different lower or upper bounds on each of the align:start position:0% lower or upper bounds on each of the align:start position:0% lower or upper bounds on each of the other parameters as well as align:start position:0% other parameters as well as align:start position:0% other parameters as well as desires to sort of hit either targets or align:start position:0% desires to sort of hit either targets or align:start position:0% desires to sort of hit either targets or make those bigger or smaller align:start position:0% make those bigger or smaller align:start position:0% make those bigger or smaller now they don't really describe very much align:start position:0% now they don't really describe very much align:start position:0% now they don't really describe very much what their align:start position:0% what their align:start position:0% what their what their align:start position:0% what their align:start position:0% what their optimization procedure is align:start position:0% optimization procedure is align:start position:0% optimization procedure is all they do is report that using the align:start position:0% all they do is report that using the align:start position:0% all they do is report that using the model they were able to achieve the align:start position:0% model they were able to achieve the align:start position:0% model they were able to achieve the following align:start position:0% following align:start position:0% following best align:start position:0% best align:start position:0% best condition which they report and are very align:start position:0% condition which they report and are very align:start position:0% condition which they report and are very satisfied with so things like you know align:start position:0% satisfied with so things like you know align:start position:0% satisfied with so things like you know very high 99 step coverage align:start position:0% very high 99 step coverage align:start position:0% very high 99 step coverage improved or fairly good align:start position:0% align:start position:0% conversion rate and so on so that's align:start position:0% conversion rate and so on so that's align:start position:0% conversion rate and so on so that's actually one of the fun things one could align:start position:0% actually one of the fun things one could align:start position:0% actually one of the fun things one could play with i think with the actual data align:start position:0% play with i think with the actual data align:start position:0% play with i think with the actual data is apply some of the hill climbing or align:start position:0% is apply some of the hill climbing or align:start position:0% is apply some of the hill climbing or other kinds of optimization algorithms align:start position:0% other kinds of optimization algorithms align:start position:0% other kinds of optimization algorithms to this data align:start position:0% to this data align:start position:0% to this data they do note in the paper that some of align:start position:0% they do note in the paper that some of align:start position:0% they do note in the paper that some of these constraints or limits align:start position:0% these constraints or limits align:start position:0% these constraints or limits do not greatly reduce their factor space align:start position:0% do not greatly reduce their factor space align:start position:0% do not greatly reduce their factor space they do not constrain align:start position:0% they do not constrain align:start position:0% they do not constrain where they need to operate and that align:start position:0% where they need to operate and that align:start position:0% where they need to operate and that gives them lots of latitude to be able align:start position:0% gives them lots of latitude to be able align:start position:0% gives them lots of latitude to be able to that's a good thing that gives them align:start position:0% to that's a good thing that gives them align:start position:0% to that's a good thing that gives them latitude to play with those parameters align:start position:0% latitude to play with those parameters align:start position:0% latitude to play with those parameters to achieve perhaps some other uh align:start position:0% to achieve perhaps some other uh align:start position:0% to achieve perhaps some other uh some other goals align:start position:0% some other goals align:start position:0% some other goals okay so what i want to do in the last align:start position:0% okay so what i want to do in the last align:start position:0% okay so what i want to do in the last six minutes is switch over here align:start position:0% six minutes is switch over here align:start position:0% six minutes is switch over here uh align:start position:0% align:start position:0% so are we ready to do a switch because align:start position:0% so are we ready to do a switch because align:start position:0% so are we ready to do a switch because what i want to do is align:start position:0% align:start position:0% okay align:start position:0% align:start position:0% okay align:start position:0% align:start position:0% okay align:start position:0% align:start position:0% so what i'm going to be doing here is align:start position:0% so what i'm going to be doing here is align:start position:0% so what i'm going to be doing here is basically showing you some jump align:start position:0% basically showing you some jump align:start position:0% basically showing you some jump analysis align:start position:0% analysis align:start position:0% analysis of this data align:start position:0% align:start position:0% yep align:start position:0% align:start position:0% oh dear align:start position:0% oh dear align:start position:0% oh dear maybe i'm not online align:start position:0% align:start position:0% so what i've basically done is just align:start position:0% so what i've basically done is just align:start position:0% so what i've basically done is just sucked in the raw data and their their align:start position:0% sucked in the raw data and their their align:start position:0% sucked in the raw data and their their input conditions and their output align:start position:0% input conditions and their output align:start position:0% input conditions and their output conditions and we'll play with two align:start position:0% conditions and we'll play with two align:start position:0% conditions and we'll play with two things i just want to give you a little align:start position:0% things i just want to give you a little align:start position:0% things i just want to give you a little bit of a glimpse here align:start position:0% bit of a glimpse here align:start position:0% bit of a glimpse here of align:start position:0% of align:start position:0% of okay align:start position:0% okay align:start position:0% okay excellent align:start position:0% align:start position:0% okay can you see that in singapore align:start position:0% okay can you see that in singapore align:start position:0% okay can you see that in singapore excellent all right i'm psyched align:start position:0% excellent all right i'm psyched align:start position:0% excellent all right i'm psyched okay so what i've done here is use jump align:start position:0% okay so what i've done here is use jump align:start position:0% okay so what i've done here is use jump jump seven in fact which is a more align:start position:0% jump seven in fact which is a more align:start position:0% jump seven in fact which is a more recent version that i'm used to so some align:start position:0% recent version that i'm used to so some align:start position:0% recent version that i'm used to so some of the controls are a little bit align:start position:0% of the controls are a little bit align:start position:0% of the controls are a little bit different uh by the way i don't know if align:start position:0% different uh by the way i don't know if align:start position:0% different uh by the way i don't know if you've played with jump before align:start position:0% you've played with jump before align:start position:0% you've played with jump before as you're looking at align:start position:0% as you're looking at align:start position:0% as you're looking at your projects align:start position:0% your projects align:start position:0% your projects uh there is a 30-day free trial license align:start position:0% uh there is a 30-day free trial license align:start position:0% uh there is a 30-day free trial license of the full version of jump so you can align:start position:0% of the full version of jump so you can align:start position:0% of the full version of jump so you can download align:start position:0% download align:start position:0% download uh the trial version if you want if you align:start position:0% uh the trial version if you want if you align:start position:0% uh the trial version if you want if you want to grab it and and play with it uh align:start position:0% want to grab it and and play with it uh align:start position:0% want to grab it and and play with it uh especially for the align:start position:0% align:start position:0% for the project i believe there's also align:start position:0% for the project i believe there's also align:start position:0% for the project i believe there's also an academic version it's like fifty align:start position:0% an academic version it's like fifty align:start position:0% an academic version it's like fifty dollars for a year license or something align:start position:0% dollars for a year license or something align:start position:0% dollars for a year license or something which is pretty nice because the align:start position:0% which is pretty nice because the align:start position:0% which is pretty nice because the commercial version is align:start position:0% commercial version is align:start position:0% commercial version is hundreds of dollars so so it's a nice align:start position:0% hundreds of dollars so so it's a nice align:start position:0% hundreds of dollars so so it's a nice package so all i did is i sucked in the align:start position:0% package so all i did is i sucked in the align:start position:0% package so all i did is i sucked in the the inputs these are our four input align:start position:0% the inputs these are our four input align:start position:0% the inputs these are our four input conditions here and then our growth rate align:start position:0% conditions here and then our growth rate align:start position:0% conditions here and then our growth rate and output conditions and i would always align:start position:0% and output conditions and i would always align:start position:0% and output conditions and i would always encourage you as a first step to explore align:start position:0% encourage you as a first step to explore align:start position:0% encourage you as a first step to explore your data just look at it see if it align:start position:0% your data just look at it see if it align:start position:0% your data just look at it see if it makes sense do univariate plots of align:start position:0% makes sense do univariate plots of align:start position:0% makes sense do univariate plots of outputs as a function of inputs and by align:start position:0% outputs as a function of inputs and by align:start position:0% outputs as a function of inputs and by the way there's some very cool things align:start position:0% the way there's some very cool things align:start position:0% the way there's some very cool things like align:start position:0% like align:start position:0% like scatter plot 3d where you might take align:start position:0% scatter plot 3d where you might take align:start position:0% scatter plot 3d where you might take three of these parameters align:start position:0% three of these parameters align:start position:0% three of these parameters and plot them against each other align:start position:0% and plot them against each other align:start position:0% and plot them against each other and so you get align:start position:0% and so you get align:start position:0% and so you get cool things like this that give you align:start position:0% cool things like this that give you align:start position:0% cool things like this that give you a picture of the scatter plot align:start position:0% a picture of the scatter plot align:start position:0% a picture of the scatter plot as a function align:start position:0% as a function align:start position:0% as a function you know of those three inputs or align:start position:0% you know of those three inputs or align:start position:0% you know of those three inputs or i guess this is a scatter plot of three align:start position:0% i guess this is a scatter plot of three align:start position:0% i guess this is a scatter plot of three outputs versus each other see if there's align:start position:0% outputs versus each other see if there's align:start position:0% outputs versus each other see if there's some correlation between these outputs align:start position:0% some correlation between these outputs align:start position:0% some correlation between these outputs you can do it one output as a function align:start position:0% you can do it one output as a function align:start position:0% you can do it one output as a function of two others um and so on so one of the align:start position:0% of two others um and so on so one of the align:start position:0% of two others um and so on so one of the nice things i like about align:start position:0% nice things i like about align:start position:0% nice things i like about jump and you should look for in a good align:start position:0% jump and you should look for in a good align:start position:0% jump and you should look for in a good interactive package is that interaction align:start position:0% interactive package is that interaction align:start position:0% interactive package is that interaction you can actually interact with the data align:start position:0% you can actually interact with the data align:start position:0% you can actually interact with the data and one of the things that excel is not align:start position:0% and one of the things that excel is not align:start position:0% and one of the things that excel is not good at so align:start position:0% good at so align:start position:0% good at so other tools like matlab are good at this align:start position:0% other tools like matlab are good at this align:start position:0% other tools like matlab are good at this as well align:start position:0% as well align:start position:0% as well okay so having done that align:start position:0% okay so having done that align:start position:0% okay so having done that another thing i like about jump is it align:start position:0% another thing i like about jump is it align:start position:0% another thing i like about jump is it allows you to align:start position:0% allows you to align:start position:0% allows you to you know align:start position:0% you know align:start position:0% you know highlight what are bad data points so i align:start position:0% highlight what are bad data points so i align:start position:0% highlight what are bad data points so i don't i can deal with missing data align:start position:0% don't i can deal with missing data align:start position:0% don't i can deal with missing data points and jump so i've pulled those align:start position:0% points and jump so i've pulled those align:start position:0% points and jump so i've pulled those from the uh from the paper as well align:start position:0% from the uh from the paper as well align:start position:0% from the uh from the paper as well and then i can start to do things like align:start position:0% and then i can start to do things like align:start position:0% and then i can start to do things like build regression models so there are align:start position:0% build regression models so there are align:start position:0% build regression models so there are graph align:start position:0% graph align:start position:0% graph oops align:start position:0% oops align:start position:0% oops i'm sorry align:start position:0% i'm sorry align:start position:0% i'm sorry a analyze align:start position:0% a analyze align:start position:0% a analyze model align:start position:0% model align:start position:0% model so here i might say okay i want to build align:start position:0% so here i might say okay i want to build align:start position:0% so here i might say okay i want to build a model for align:start position:0% a model for align:start position:0% a model for i'll just pick one for the moment growth align:start position:0% i'll just pick one for the moment growth align:start position:0% i'll just pick one for the moment growth rate as a function of these four input align:start position:0% rate as a function of these four input align:start position:0% rate as a function of these four input parameters and i can do things like say align:start position:0% parameters and i can do things like say align:start position:0% parameters and i can do things like say i really am doing a response surface align:start position:0% i really am doing a response surface align:start position:0% i really am doing a response surface model up to polynomial degree 2 align:start position:0% model up to polynomial degree 2 align:start position:0% model up to polynomial degree 2 and it builds for me align:start position:0% and it builds for me align:start position:0% and it builds for me the template if you will for all of the align:start position:0% the template if you will for all of the align:start position:0% the template if you will for all of the interaction terms the square terms all align:start position:0% interaction terms the square terms all align:start position:0% interaction terms the square terms all of those things that if you were doing align:start position:0% of those things that if you were doing align:start position:0% of those things that if you were doing this in excel you would have to you know align:start position:0% this in excel you would have to you know align:start position:0% this in excel you would have to you know yourself build those fictional align:start position:0% yourself build those fictional align:start position:0% yourself build those fictional input parameters but here i'm basically align:start position:0% input parameters but here i'm basically align:start position:0% input parameters but here i'm basically telling it directly what the structure align:start position:0% telling it directly what the structure align:start position:0% telling it directly what the structure of the model i want to fit is align:start position:0% of the model i want to fit is align:start position:0% of the model i want to fit is now one thing that we haven't talked a align:start position:0% now one thing that we haven't talked a align:start position:0% now one thing that we haven't talked a lot about that i wanted to to highlight align:start position:0% lot about that i wanted to to highlight align:start position:0% lot about that i wanted to to highlight is align:start position:0% is align:start position:0% is how to go about building the model i align:start position:0% how to go about building the model i align:start position:0% how to go about building the model i could go ahead and do a full anova a align:start position:0% could go ahead and do a full anova a align:start position:0% could go ahead and do a full anova a full regression and a nova and look at align:start position:0% full regression and a nova and look at align:start position:0% full regression and a nova and look at each of the terms and decide what's align:start position:0% each of the terms and decide what's align:start position:0% each of the terms and decide what's significant and what's not align:start position:0% significant and what's not align:start position:0% significant and what's not there is a align:start position:0% there is a align:start position:0% there is a stepwise align:start position:0% stepwise align:start position:0% stepwise version of that same process align:start position:0% version of that same process align:start position:0% version of that same process where align:start position:0% where align:start position:0% where essentially align:start position:0% essentially align:start position:0% essentially what we do is one parameter at a time align:start position:0% what we do is one parameter at a time align:start position:0% what we do is one parameter at a time add align:start position:0% align:start position:0% i wonder if i lost my align:start position:0% align:start position:0% okay can you see can you see it uh just align:start position:0% okay can you see can you see it uh just align:start position:0% okay can you see can you see it uh just uh maybe we'll just transmit i'm i'll align:start position:0% uh maybe we'll just transmit i'm i'll align:start position:0% uh maybe we'll just transmit i'm i'll i'll try align:start position:0% i'll try align:start position:0% i'll try it's not accepting the net meeting align:start position:0% align:start position:0% yeah it won't look as good but let's align:start position:0% yeah it won't look as good but let's align:start position:0% yeah it won't look as good but let's let's at least do that since we only align:start position:0% let's at least do that since we only align:start position:0% let's at least do that since we only have a couple more minutes align:start position:0% align:start position:0% yeah align:start position:0% yeah align:start position:0% yeah yeah that's fine align:start position:0% yeah that's fine align:start position:0% yeah that's fine so we're just going to project so you align:start position:0% so we're just going to project so you align:start position:0% so we're just going to project so you can see see kind of crudely on the align:start position:0% can see see kind of crudely on the align:start position:0% can see see kind of crudely on the screen but the basic idea of this align:start position:0% screen but the basic idea of this align:start position:0% screen but the basic idea of this stepwise regression is that you can align:start position:0% stepwise regression is that you can align:start position:0% stepwise regression is that you can set the significance level for accepting align:start position:0% set the significance level for accepting align:start position:0% set the significance level for accepting a model parameter align:start position:0% a model parameter align:start position:0% a model parameter and then one step at a time so for align:start position:0% and then one step at a time so for align:start position:0% and then one step at a time so for example if i do this step align:start position:0% example if i do this step align:start position:0% example if i do this step it's it's pulling in align:start position:0% it's it's pulling in align:start position:0% it's it's pulling in additional coefficients that are of align:start position:0% additional coefficients that are of align:start position:0% additional coefficients that are of acceptable significance and in fact i align:start position:0% acceptable significance and in fact i align:start position:0% acceptable significance and in fact i think the probability to enter it needs align:start position:0% think the probability to enter it needs align:start position:0% think the probability to enter it needs to be align:start position:0% to be align:start position:0% to be it's set to 0.25 this is 75 confidence align:start position:0% it's set to 0.25 this is 75 confidence align:start position:0% it's set to 0.25 this is 75 confidence level needs to be achieved in order to align:start position:0% level needs to be achieved in order to align:start position:0% level needs to be achieved in order to accept align:start position:0% accept align:start position:0% accept the model and i can keep stepping align:start position:0% the model and i can keep stepping align:start position:0% the model and i can keep stepping through and it iteratively is basically align:start position:0% through and it iteratively is basically align:start position:0% through and it iteratively is basically making determinations based on align:start position:0% making determinations based on align:start position:0% making determinations based on my setting of significance values to align:start position:0% my setting of significance values to align:start position:0% my setting of significance values to pull in or discard model terms align:start position:0% pull in or discard model terms align:start position:0% pull in or discard model terms and if it discards a model term then it align:start position:0% and if it discards a model term then it align:start position:0% and if it discards a model term then it takes those degrees of freedom throws align:start position:0% takes those degrees of freedom throws align:start position:0% takes those degrees of freedom throws them back into the error term align:start position:0% them back into the error term align:start position:0% them back into the error term not purely as a function of you know the align:start position:0% not purely as a function of you know the align:start position:0% not purely as a function of you know the replicate error but it gives it it does align:start position:0% replicate error but it gives it it does align:start position:0% replicate error but it gives it it does this kind of stepwise align:start position:0% this kind of stepwise align:start position:0% this kind of stepwise multiple regression in order to build align:start position:0% multiple regression in order to build align:start position:0% multiple regression in order to build the overall model align:start position:0% the overall model align:start position:0% the overall model so here you can see for example for the align:start position:0% so here you can see for example for the align:start position:0% so here you can see for example for the this particular output of growth rate align:start position:0% this particular output of growth rate align:start position:0% this particular output of growth rate you can see which terms end up being align:start position:0% you can see which terms end up being align:start position:0% you can see which terms end up being significant you've got things like align:start position:0% significant you've got things like align:start position:0% significant you've got things like temperature squared align:start position:0% temperature squared align:start position:0% temperature squared and temperature h2 pressure interactions align:start position:0% and temperature h2 pressure interactions align:start position:0% and temperature h2 pressure interactions okay doing that you can then accept one align:start position:0% okay doing that you can then accept one align:start position:0% okay doing that you can then accept one of these models and build another column align:start position:0% of these models and build another column align:start position:0% of these models and build another column in align:start position:0% align:start position:0% in jump that is in fact the prediction align:start position:0% in jump that is in fact the prediction align:start position:0% in jump that is in fact the prediction formula for one of these parameters so align:start position:0% formula for one of these parameters so align:start position:0% formula for one of these parameters so for example this might be the prediction align:start position:0% for example this might be the prediction align:start position:0% for example this might be the prediction formula for growth rate it now has that align:start position:0% formula for growth rate it now has that align:start position:0% formula for growth rate it now has that polynomial model in it and one of the align:start position:0% polynomial model in it and one of the align:start position:0% polynomial model in it and one of the cool things about jump just so that you align:start position:0% cool things about jump just so that you align:start position:0% cool things about jump just so that you know it's there and then you can play align:start position:0% know it's there and then you can play align:start position:0% know it's there and then you can play with it align:start position:0% with it align:start position:0% with it or you can also now start to go in and align:start position:0% or you can also now start to go in and align:start position:0% or you can also now start to go in and do things like align:start position:0% do things like align:start position:0% do things like profiler or contour profiler for align:start position:0% profiler or contour profiler for align:start position:0% profiler or contour profiler for optimization and exploration of that align:start position:0% optimization and exploration of that align:start position:0% optimization and exploration of that design space align:start position:0% design space align:start position:0% design space so doing a a simple align:start position:0% so doing a a simple align:start position:0% so doing a a simple one parameter profiler i might say okay align:start position:0% one parameter profiler i might say okay align:start position:0% one parameter profiler i might say okay i want to profile growth rate and row align:start position:0% i want to profile growth rate and row align:start position:0% i want to profile growth rate and row i'll just do those two align:start position:0% i'll just do those two align:start position:0% i'll just do those two parameters align:start position:0% parameters align:start position:0% parameters and what i've got is this prediction align:start position:0% and what i've got is this prediction align:start position:0% and what i've got is this prediction profiler that has align:start position:0% profiler that has align:start position:0% profiler that has the align:start position:0% the align:start position:0% the the two outputs growth rate is the first align:start position:0% the two outputs growth rate is the first align:start position:0% the two outputs growth rate is the first row and align:start position:0% row and align:start position:0% row and whatever it was align:start position:0% whatever it was align:start position:0% whatever it was is it i can't remember what the second align:start position:0% is it i can't remember what the second align:start position:0% is it i can't remember what the second one was it's a little hard to see there align:start position:0% one was it's a little hard to see there align:start position:0% one was it's a little hard to see there formula row okay so the resistance and align:start position:0% formula row okay so the resistance and align:start position:0% formula row okay so the resistance and the growth rate as a function of each of align:start position:0% the growth rate as a function of each of align:start position:0% the growth rate as a function of each of the individual parameters so this is the align:start position:0% the individual parameters so this is the align:start position:0% the individual parameters so this is the univariate dependence depending on my align:start position:0% univariate dependence depending on my align:start position:0% univariate dependence depending on my current operating point align:start position:0% current operating point align:start position:0% current operating point and then i can move that operating point align:start position:0% and then i can move that operating point align:start position:0% and then i can move that operating point and see how the other parameters align:start position:0% and see how the other parameters align:start position:0% and see how the other parameters change align:start position:0% change align:start position:0% change so as i change temperature align:start position:0% so as i change temperature align:start position:0% so as i change temperature how does that affect my my outputs align:start position:0% how does that affect my my outputs align:start position:0% how does that affect my my outputs okay so it's kind of fun you can play align:start position:0% okay so it's kind of fun you can play align:start position:0% okay so it's kind of fun you can play around align:start position:0% around align:start position:0% around and see how align:start position:0% and see how align:start position:0% and see how how those things happen you can also do align:start position:0% how those things happen you can also do align:start position:0% how those things happen you can also do essentially close to what they did in align:start position:0% essentially close to what they did in align:start position:0% essentially close to what they did in the paper and do contour plot align:start position:0% the paper and do contour plot align:start position:0% the paper and do contour plot explorations so i might for example do align:start position:0% explorations so i might for example do align:start position:0% explorations so i might for example do those same parameters align:start position:0% align:start position:0% yeah that's fine align:start position:0% yeah that's fine align:start position:0% yeah that's fine and actually now plot the current align:start position:0% and actually now plot the current align:start position:0% and actually now plot the current operating point the response align:start position:0% operating point the response align:start position:0% operating point the response of align:start position:0% of align:start position:0% of growth rate row align:start position:0% growth rate row align:start position:0% growth rate row resistance uniformity align:start position:0% resistance uniformity align:start position:0% resistance uniformity in the design space so that i can see align:start position:0% in the design space so that i can see align:start position:0% in the design space so that i can see how those things change as i change my align:start position:0% how those things change as i change my align:start position:0% how those things change as i change my gap spacing or my temperature and i can align:start position:0% gap spacing or my temperature and i can align:start position:0% gap spacing or my temperature and i can also apply things like the limits maybe align:start position:0% also apply things like the limits maybe align:start position:0% also apply things like the limits maybe i have a low limit of 500 for the growth align:start position:0% i have a low limit of 500 for the growth align:start position:0% i have a low limit of 500 for the growth rate i need again 500 nanometers and align:start position:0% rate i need again 500 nanometers and align:start position:0% rate i need again 500 nanometers and what it does is it'll start to shade out align:start position:0% what it does is it'll start to shade out align:start position:0% what it does is it'll start to shade out unallowable regions to achieve those align:start position:0% unallowable regions to achieve those align:start position:0% unallowable regions to achieve those kinds of goals align:start position:0% kinds of goals align:start position:0% kinds of goals so this is just a nice interactive tool align:start position:0% so this is just a nice interactive tool align:start position:0% so this is just a nice interactive tool that lets you do that kind of align:start position:0% that lets you do that kind of align:start position:0% that lets you do that kind of uh dynamic exploration of the optimism align:start position:0% uh dynamic exploration of the optimism align:start position:0% uh dynamic exploration of the optimism of the optimal space align:start position:0% of the optimal space align:start position:0% of the optimal space so i'll put the uh put that data uh up align:start position:0% so i'll put the uh put that data uh up align:start position:0% so i'll put the uh put that data uh up on the align:start position:0% on the align:start position:0% on the the web as well and you can play around align:start position:0% the web as well and you can play around align:start position:0% the web as well and you can play around with it but this is uh uh kind of where align:start position:0% with it but this is uh uh kind of where align:start position:0% with it but this is uh uh kind of where i'll i'll leave it for you to explore align:start position:0% i'll i'll leave it for you to explore align:start position:0% i'll i'll leave it for you to explore more if you want to with jump either on align:start position:0% more if you want to with jump either on align:start position:0% more if you want to with jump either on this data or align:start position:0% this data or align:start position:0% this data or perhaps more efficiently on your team align:start position:0% perhaps more efficiently on your team align:start position:0% perhaps more efficiently on your team project data align:start position:0% project data align:start position:0% project data but wanted to get get this case study so align:start position:0% but wanted to get get this case study so align:start position:0% but wanted to get get this case study so you know some of the capabilities so i align:start position:0% you know some of the capabilities so i align:start position:0% you know some of the capabilities so i think with that i'm going to break align:start position:0% think with that i'm going to break align:start position:0% think with that i'm going to break and we can chat a little bit if people align:start position:0% and we can chat a little bit if people align:start position:0% and we can chat a little bit if people have questions about team projects and align:start position:0% have questions about team projects and align:start position:0% have questions about team projects and and whatnot