 Okay, here we go, till now we have seen many processes and also we have seen how to actually fabricate an IC and how many process steps we go through. The simplest process which I have shown last few days, 16 mass process and you can imagine now any additional feature I add, I will have another mask on it, another metal, another mask and if it is a tungsten stub I may have another mask for it. So there are many masks keep adding and each mask cost a million dollars plus. So it is not that so trivial that I add another mask. So process wise one has to plan very carefully how many masks will be required and that is a designer's job as well as process people's job. So now some way now in 2012 onwards or 10 onwards both technology and designers should sit together and actually plan what is the system to be designed. So it is not arbitrary earlier in 1970's a very famous professor, Carver Mead at Caltech, he suggested and which he successfully did that as well that a computer scientist can design a chip without knowing absolutely anything of technology. He himself was a professor of computer science, he was a professor in electrical engineering, he was a professor in applied sciences, material sciences and he was, there is called Currency Appointment, he was in 7 departments in spite of all of that and this was his first book on VLSR design appeared. This is the first textbook in VLSR design appeared in 78 from Carver Mead. Of course there is another author with him was Ms. Lynn Conway, she was with Xerox companies and surprisingly the first IC chip interest was shown by Xerox company and not by any Intel or anyone you know first such chip usefulness was shown by Xerox people and Lynn Conway was the chief of design group in Xerox okay. So this is some history, so I just want to tell you that now the technology knowledge is as much needed for designers as is designs knowledge is as much needed by technology people and together should also know little bit of physics, little bit of chemistry, material science, optics and whatever whatever you think about okay, little bit at least you should at least terminology should be known what is it talking about okay. Okay so we start now with the new topic and we have seen in that last so many days that certain films we grow but there are very few things we can grow mostly silicon dioxide can be grown out of silicon. But the rest of the time I will have to keep depositing films of different kinds. So the process which is very dominant in all IC manufacture is depositions and anything you deposit we see we have seen other day that there will be selective etching of something so you need to etch something so the next part of this is etching because how good etching you do will be deciding the characteristics. So etching and deposition to some extent are not necessarily exactly complementary but to some extent they are complementary whatever you deposit somewhere you etch okay. So somewhere you etch means somewhere you deposit so there is some theory at least some processes you can say like RF depositions what we do normally RF based they can be used by just changing the RF source polarities as a nature okay so that is exactly what you do so that is why I am trying to say you that many things are similar identical but not everything so let us see each of them separately today we start with a little basic of thin film depositions and at least two of the most common methods of depositing films will discuss and unfortunately all that I will discuss today is last rarely used these days in manufacture okay but they are the precursors our labs have many such evaporators so I think we will discuss that in an integrated circuit fab many films of different materials are deposited typically films can be metals metal silicites they can be dielectrics like nitride oxide many others and so these are all films you need to deposit so a typical system which allows you to deposit and the film which we are going to get this is essentially called film deposition because normally the thickness of these depositing films are thickness of deposited layers are very thin comparatively it is not 300 microns or 10 microns it may be few thousand Amstrons or 2000 Amstrons or at best half a micron 5000 maximum so in that case they are called films okay films are actually we are not doing this course here but otherwise I mean tell you there is a another technology which is used in SOC the SICs now SIPs is called thin film techniques and thick film techniques okay so the idea there was thick film techniques will be thicker films and thin film techniques will be thinner films there are requirements when thin films are thicker than thicker films and thick films are thinner than thinner films so all games are only namesake but let us look at the requirements of deposition any system which I use should have it should be able to give me desired composition of film which I want to deposit should not dissociate if I am putting silicon dioxide it should retain silicon dioxide property after depositions and it should not allow any contaminants with it because if it allows contaminant then whole process will get tarnished and we will not be able to proceed further so lay in contaminants and any composition should be retained after the deposition process is over okay most cases if they are used in silicon IC manufacturers as interconnects or many other areas or even insulating films they should actually retain their electrical property in the best fashion if you are a conductor it should give largest conductivity possible which should not modify its conductivity if you are a dielectric film it should retain its dielectric constant okay and also porosity so one of the major requirement of any system is that it should remain good electrical property the film which are desired by you for the next processings the mechanical properties should also be very good there is some very important property maybe I will show you later when you deposit a film you have a molecules or atoms coming mostly atoms in some system molecules also can come so these atoms are impinging on substrate they should take other way move away okay so there should be something called addition or sticking coefficient available unless there is a sticking these atom please remember what is the problem these whatever this energy or whatever this the ions electron I am sorry the films atoms are coming they are energetic they are not zero energies so when they say they lose their kinetic energy and they may actually move okay so worries are that that energy should not be so high that they actually display something and move themselves so it should be retained as the film they want we want to that is very typical so a vision is very important of course we will see in CVD ladder there are two kinds of a vision properties or vision energies we use one is called surface energy the other is called volume energy so when to have what higher or lower we will decide whether film will grow or it will etch okay so a mechanical property is very crucial for any deposition system it should retain and give whatever property I am looking for then I want any film I am depositing on a wafer of whatever size I may have 4 inch 8 inch 12 inch 16 inch wafers it should be uniformly deposited I want the same thickness same compositions that is minimum what I am expecting very favourable changes then I do not know for each chip or each area what the film thickness okay so that is it should wafer to wafer on the sorry on wafer wafer to wafer and run to run their thickness should be as what you are looking for the next important process step next important requirement is as we have seen yesterday vias for example if there is a step something like this so the metal film or whatever film you are depositing it also should follow that step okay coverage has to be full what it can do is if this is my step I may actually get this in some regions there may not be any sticking okay so there has to be a property of your system which will allow coverage of all steps conformality is the most required property in any IC manufacture okay so this is as I say step coverage is the most important part which system should allow and the material should go inside that wherever vias or voids are created if inter if the films are metal and they are used as interconnects as I just now said they should provide you very low resistance per micron or per length okay because essentially if you are running an interconnect what do you mean by interconnect some signal is going from one point to the other okay and the length of interconnect will be large enough compare that is why I call interconnects this R is larger what will it create problem the delay RC time constant will increase and it may happen that this RC time constant equivalent frequency may be same as what your system clock or your signal frequency is going so you may actually have 180 degree out of phase signal going so it is a very typical problems one going to 0 0 becoming 1 because of delay signal because you most circuits are these days there are 2 kinds of circuit is preserved this I talked about one is gale okay think of it there is something some circuit has to be locally synchronous globally asynchronous some circuits are globally asynchronous locally synchronous synchronous means time with the clock which you generate or you have in your system okay so all signal need not be there may be some asynchronous signals going and many process many such systems do use asynchronous transfers okay it does not mean please also for the circuit people I may say asynchronous does not mean there is no clock okay it is not a static logic asynchronous essentially may it is not running at the system clock is that word clear so asynchronous are nothing to do with non clocks okay it may still be clock system for example one of the feature which I keep asking the m tech student when they come for interview take a latch to inverter's back to back there is absolutely no clock why it should be called asynchronous system it is a time system when there is no clock okay so how it acts as a clock when there is no no clock going given to it but a latch is essentially a time signal based system okay similar thinking has to be done in designs so some are asynchronous some are synchronous some are non-looking synchronous okay latches for example so these are issues are independent of all that but this RC time constraints since it affects that very much of course one method which circuit people use is buffer it keep pumping every after certain length you actually have a buffer stage and you pump it again so the delay is taken care through the buffer out however that means you need any additional power every certain length because your signal pumping every now and then this additional power is not good because you are reducing the power of the system and you are saying I'll put buffers it will take area large area because it will larger buffer so all these issues are related to interconnects so to some extent my giving a correct interconnect is a crucial part for me but it will add then affect the their performance of the speed speed of a circuit so somewhere don't think that interconnects are very true interconnects technology is very crucial because that decide the performance at the end of the day okay all this I just want to show you how around we think same way this interconnect have a problem which is called electro migration typically electro migration can be thought like this you have some substrate or something and you are a metal layer okay now this metal layer are getting thinner and thinner okay because of scaling also we are finding it now that this film which we are this earlier our since we are not scaling voltages the currents in because area is now reducing current density which wire is now taking or thin is taking is increasing that here the number of electrons per square centimeters are now more okay compared to earlier one because your thicker of this the per unit area will be smaller now it is larger current densities if the current density is very high which can be typically say greater than equal to 10 to power 5 m per centimeter square per cm square then this is called critical current density C is greater than JC then there are huge number of electron densities available and ions there but since there are huge number of electrons per unit area and you are applying a voltage that is why the current flows so it creates a electric field and large number of electrons are accelerated with this electric field and they create a effect which is called wind effect okay they create a effect ions are very immobile thing they are very heavy they stick to where they are but electrons are very light they are acquiring energy because of large number of carriers per cc per centimeter square and we are applying currents voltages they actually pick up enough energy from the fields okay these create something called wind effect and they have enough force kinetic energy acquired to actually dislodge ions okay as actually dislodge ions so if there is a small crack in your metal film groups wherever it is somewhere here so this whole metal will actually climb from here and there will be a gap between the left side to the right side the metal will climb up okay from that cracks so there will be a gap between interconnect and when this will happen it may not happen in first date the crack may respond initially so initially circuit may work after some day you find it did not work everything was connected everything was fine chip was working and tomorrow we used to say in class it is electronic mood the mood is migration okay the metal actually climbs okay so there is a reliability issue this is called reliability aspect and one has to worry that any time and that is why someone was asking other day copper intercopper has better migration compared to aluminium aluminium as the lowest migration density so yet even going to power 4 or little higher it may migrate okay so this immigration is a major issue and probably even now in all SOC systems or all systems which are hardware so co-ware designs people are doing one of the major worries is migration volume is not going to be that but area you are reducing all the time and fields you are not reducing increasing because voltage you are not scaling if you scale everything exactly nothing should happen but it is not happening so there are issues which are called reliability issues which is hurting both process people so one method when we when we actually lay the interconnect there is a case there itself we should think if this is going to carry how much carrier how much currents in this area we know by because I am designing chip I should actually add additional features there thicker I must announce I need thicker or at least wider okay so all this interconnect design has becoming very very crucial now as we are scaling down in 90s when I had many student and B I myself used to design now I do not those days this was all relevant 5 micron 3 micron 1 micron all fine everything is fine things are changing so fast now that every day one has to worry about failures and failure means money so all the problem is economics so please remember this retro migration is not real now people are looking from layout people say you will look at the layout and declare how much yield it can give okay so there is a cat tools have been developed for layout generation to rely on issues so these are all problems of what you all think new problems for 2014 the next problem which a metal system or any film system I am expecting is non-corrosion it should not oxidize particularly copper film for example it should not become copper oxide and please remember copper oxide is not just one kind of copper copper has two balances and can be C2O or COO COO and both are as bad as corrosion is concerned some are lesser etchability others have stronger etchability but depending on how much oxygen it picks up I may have worse phases okay so this is very crucial in actual processing then of course this you can write whenever someone asked desired this this line you can without thinking right economical what costly nahi chahi that is why we do not use gold so often gold is actually not as costly as platinum so we use platinum and keep saying gold is costly so this is something another if you go to a jewellery you never buy platinum of course nowadays I am told they I do not have enough money to buy so I do not know but otherwise platinum jewellery is also available which is costly as a gold jewellery but gold has that aura platinum does not have in the lab if I have a platinum target I may even keep it like this no one will pick up but if I have a gold target I am sure next day it will not be there if I just keep it okay so the worries are terrific in economics okay and all similarly the 10th also you can always say it should be compatible to all other processes because suddenly you say 1200 degree by current no no no no we have 800 fixed 1000 fixed no one more than that whatever compatible other processes are this cannot be different from okay so these are I mean these are all not that it is not given in a book but I just numbered you to show that what is so much important for a deposition system and depositing materials there are two major methods of deposition of films one is called physical vapor deposition the other is called chemical vapor deposition okay first we will only do today possibly I am not sure but possibly PVDs which is physical vapor deposition and as I keep saying the A there are two kinds of physical vapor deposition so one is called evaporation the other is called sputtering okay other is called sputtering this I am not sure whether we will finish today because it needs some plasma so maybe we will do something on plasmas today the second possible mechanism is chemical vapor so you actually have a source which can be of any form but should convert to gaseous form and those whatever material you want to deposit the gas form of that should move on the substrate and by some temperature pressure adjustment it should stick to the surface and deposit okay is that right the gas will come it will some way connect to the surface of the substrate which you are kept at a given temperature and pressure there is some mathematics we can create and we say when it will stick okay so a layer of that species can deposit on it can be silicon it can be silicon dioxide it can be silicon noitide or for that matter any metal okay so CVD is very popular but as I said there are number of CVD is possible one is called atmospheric pressure CVD low pressure CVD cos minus CVD and hot wire CVD if you are working with Professor Ram Gopal Rao he will take you someday to ground floor lab inside and see that you know this is hot wire CVD everything can be done there okay let us do evaporation first and once we do evaporation we will say okay there are some problems evaporation and as I said no one is now want to work with that okay so we went for sputtering which is another PVD sputtering is still used 100% used but it has also has its own problems okay like for example this conformal depositions are very difficult in sputtering systems so it can deposit it is not that it cannot you need some special gadgets to do that whereas CVD normally will always go through conformal or contours it will take steps okay so in most cases where and all ICs we have seen so far every step there is a different thicknesses everywhere so any film goes it has to actually follow everywhere okay and that is our worry actually if it does not follow then we have a worry if it follows also we have a worry because then we have to do CMP we want to planarize so there are all issues okay so we will look into the first one evaporation PVD actually both PVD and sputtering has some basic similar but only method of this is also deposition by kind of evaporation sputtering but it is slightly different in nature so we will discuss it later. Typical PVD process has 4 possible steps or looks like that you may have a source material which you want to deposit so it may be in the solid phase or it may be in the gaseous phase we say condensed phase okay except gas both liquid and solid are called condensed matter please remember this is a if you are not knowing this do not walk into the last classes of the word corridor chemical chemistry physics they shouted you okay the condensed phase the first thing we do is by some method process which is called evaporation this condensed phase changes to gaseous phase okay now this gaseous phase is then transported to some place where we want again in the gaseous phase that means something is coming flux of gas is coming it is moving up and it is also reaching here so that is the gas phase position where it has reached and at that point which is my substrate where I want to deposit something it converts into solid phase okay so this is typically the way process of evaporation or process of any deposition takes place okay repeat starts with any of the solid or liquid source then it evaporates convert to gas phase transports still in gas phase condenses and deposits okay these are only my figures to show how things happen I could have said in one line but just now it sticks to you that this is how it happens okay why there is something I have to do here why I have to do something here okay why something I have to be done here because if I just say something you will not appreciate that each replaces I have to do something to actually do better there that is why the fourth stage that is the only thing I can add from the book which book does not know okay I do not know plumber any time visited lab but hopefully okay I will show you the list of them then you will not ask me just a minute wait for whatever things we can deposit we are answering electronic I can deposit anything anything okay so what are a typical requirement of a evaporation system is it should have a vacuum okay the system should need metal films or any films they can only go properly if you have a vacuum system all around there is a vacuum okay typical vacuum which is needed in evaporation is better than 10 to power minus 6 torr okay 760 torr is one atmospheric pressure okay now you go and look at the Google there are there is a bar there is a Pascal there are dines force per unit this centimeter square this per second per second all kinds of units are there for pressures okay just look at it how many at least 5 or 6 units are there for pressures generally because the torsely was the person who actually worked on vacuum he was trying to show vacuum as huge potential of pressure building and because of that that credit was given to him and the pressure was then called tors and less than one atmospheres are normally expressed in tors otherwise it will be in the larger this which will be either force force per unit area or it may be Pascal's or it may be bars okay 1335 Pascal's is one torr smaller units okay I give you some numbers I have my 32 years I have gone through all kinds of units so this 10 to power minus 6 torr is a very low pressure means better vacuum this world also should go better vacuum means low pressure okay I have a better vacuum so it does not mean 10 to power 1 atmospheric pressure it is not a vacuum it is a atmospheric pressure okay so I need a better vacuum possibly better than 10 to power minus 6 means minus 7 10 to power minus 8 10 to power minus 9 and even if it is better than minus 9 it will be good for something but it may be bad for something so I do not want to go below 9 in some specific molecular beam epitaxia I may go for minus 12 which I may not be able to retain but I want minus 2 but how to generate this vacuum that itself is a problem because I need vacuum to be actually constantly place there wherever my wafers are there where is my source of deposition system is everything is inside a chamber which is vacuum I actually vacuates and it should maintain the pressure why I am constantly telling maintain as things will evaporate what will happen pressure will actually increase okay so vacuum will go spoil it will spoil so I should know if I am depositing this much time so my vacuum has to be little higher better earlier so that even with this it should not go below minus 6 that is the word better as I say ultra high vacuums are better than minus 9 there are different kinds of pumps may be some I list them later which one can do to actually evacuate heating system evaporation is normally a process in which the film which you want to deposit it I just now showed the 4 stages so I want to convert from solid or liquid to the gaseous state so I must provide thermal energy kinetic energy so that it evaporates now the question arises why vacuum one of course I will show you some other reason the most important reason why I am going for vacuum is if I dip I can heat aluminum wire for example or any wire by some other ceramic heaters 1400 1800 3000 it will start operating nothing wrong with it but all oxygen around okay so it will oxidize in no time so all that I want the film to be as that is what the requirement are listed so I want that there is no oxygen content in better the vacuum the gas inside will be very very small and mostly nitrogen okay it is more inert okay so we would like to see that better vacuums are used because it will not oxidize the film during heating no oxygen around of course it does have the quartz chamber which we put releases oxygen so it is not all that trivial but still so one method is of course you heat it actually heat it okay it is called thermal heating the other is you give an energy of heat for heating by electron beams okay and we will show you the two of them typical energy which we use is 1 to 10 KV Kilo electron volt Kilo watt is the power but if the energy is KV so to keep a vacuum it should be covered some way okay so that is called bell jar and also whatever film I am growing I am depositing I should be aware that how much film thickness I am going I am depositing well if too thick a film I do not want and too thin also I do not want so I must monitor it the monitoring system should also receive the films but should not interfere with the substrate any time so thickness monitor is also a part of any system then there is another problem which one sees how do you stop the evaporation one method is you shut off the power do not evaporate but there is a still retentive heat as we call that and it does not go immediately so it will still evaporate even if you such of the power some we already heated more than desired so by the time it goes below that temperature where it evaporates it will still evaporate that means the thickness will not be what you are looking for okay now this is an issue which therefore needs some kind of stopping physically so put a shutter so as soon as my operation over I bring shuttered above the source nothing can go to my substrate it will hit to the shutter system so then therefore we need a mechanical shuttle to control the flux going to the substrates typical vacuum system is shown here which is not the best drawing or neither the best correct actual evaporating system you already earlier we used to acquire Beljahr nowadays we have stainless steel Beljahrs which is slightly better compare because why stainless steel is better very 316 in specific as a very low content of oxygen 304 has those are metallurgists should tell me why see 304 steels are stainless steels are not as good as 316 these are the numbers okay so stainless steel is not the only steel unless chromium is added there is no stainless on that okay the typical Beljahr this is where it is earlier we used to have quarts now we will have stainless steel jars if it was that is why we left quarts are because quarts releases some oxygen so we wanted to get rid of that then there is some holder over which the impurity I mean the film which you want to deposit pieces or gaseous whichever form you want to keep you can keep there okay then this red part I added just to make a shutter system which has a out please remember the trick of the trade is all its operating knobs are outside the vacuum okay but if I know anything the vacuum should not change so something has to be done some kind of Wilson seal as we call which allows us the movement without actually actually when I move it presses hard so it actually does not allow vacuum to go back so we need to have all kinds of Wilson seals around so that no vacuum is released otherwise there will just get in I move something it will air will go so all controls every control electrode everything is controlled from outside which are mostly mechanical or electromechanical but they should be separated from the jar they should get in but should not but any time I do motion it should not change my vacuums how to create a vacuum I have vacuum system which is connected to 3 or 5 there are two kinds of if you only want minus 6 or minus 7 torque when I say minus 6 mean 10 to power minus 6 or minus 7 torque then you can have only two simple pumps the first is called rotary pump in market it is called mechanical pump and the second pump is diffusion pump okay so a diffusion pump can only operate if the pressure below it is 10 to power minus 2 I mean where it is sucking diffusion pump cannot operate in atmospheres so it needs vacuum at least into power minus 2 to improve on that 10 to power minus 2 torque I can create by rotary pumps so I have a rotary pump followed by diffusion pump so first rotary pulls the vacuum to minus 2 and then I start diffusion pump and once diffusion start from starts it will start evacuating gases from above and will exhaust out this system may be some other time I can show you there are 7 8 kinds of pumps each has own problems turbo non-turbo but anyway right now this course will look into this if you are in a lab at least that is what we have been told when we were students or we were even an engineering people in TIFR that if you spoil a system you have to repair okay so if vacuum pump does not work during your evaporation it is your job to open it see what has gone wrong read it understand it repair it and fix it back because other person is waiting to be use that so that is the method we learn so we almost know 8 to Z of system which we work simply because one in so many years something will fail when I am working then I will be stuck 4 months I spent only on designing a electron beam system I have no idea how it is your one is not available in market so I designed my own okay so please I go this if you vacuum it then there is a heating something here which I can create heater which will convert this into gaseous phase I open a shutter so this gas molecules or gas atoms depends on what you are operating will actually move upwards there is a target system where substrates are held this typically this target system is some kind of spherical or cylindrical or the spherical part of the sphere okay and the wafers are held everywhere and very important thing which we do there is if in a large half sphere there may be say 24 or 32 small wafers you can fix so I have a mechanical motor on every wafer on the top so wafer actually rotates okay and this whole spherical system also rotates what is it called planetary system so in all evaporation system we have a planetary system in which whole sphere of the holder is moving and each wafer is also rotating why we did this uniformity and step coverage any angle this time when it will come I will see operation that side so step coverage as well as uniformity is possible by planetary system what is planetary system half spherical beam inside which there are wafer holders as many wafers you can hold and this whole system keeps moving and hold each wafer which has a small motor on that DC motors now they are degassing so scobay protect them they keep rotating okay there is a rate at which you should rotate okay so there are all kinds of electrical engineering goes in designing this planetary system okay now question arises few things before of course we will come back it again why vacuum one of the thing I said it that oxygen I do not want or any other species which can create some kind of impurity I do not want to happen but that is not the only reason I want vacuum there vacuum has more relevant requirement is we will see this formula little later depending on atomic radius of any species which you are operating the mean free path of this gas molecules or these gas atoms is decide is inversely proportion to pressure okay is that okay so what happens if I want let us say this distance is 1 a 1 feet 30 centimeter if I want the incoming items to reach substrate without collision why is the collide what will happen they will split everywhere else why they will not go to the substrate area I want them to go upwards without collisions the collisions means if the collisions will not happen if the distance I mean if the mean free path is larger than the distance you are traveling okay the pressure and mean free paths are inversely proportional so larger the vacuum larger is the mean free path so more likely you will get flux which is going up okay so two reasons vacuum is needed one is molding of the operation is a therefore since I will say I need to do all processes in the system I must provide you the growth rate okay or time for which this film thickness is achievable so I must create model okay for everything which I do reality in my lab I should able to create equivalent model only then computer also can do same process so here is a model for evaporation we have some definition first we define a partial pressure p star of a gas which is in equilibrium with its condensed state okay you have a condensed state solid or liquid and what is gaseous total pressure which is exerting on it is called partial pressure at a given temperature this is fixed okay so for a species so that is p star what is the total pressure which is outside is that in a chamber there is a net pressure but at the vacuum to the I mean at the material to this whatever number of gases they will also exert a pressure that is called partial pressures there is a Gary Lucet law in chemistry lead it okay what is the difference whenever there is a condensed state and you are operating or some reason the papers are there papers are always there even at room temperature but their pressures are much lower because the amount of gas which is available in room temperature is very small when you heat enough molecules are there they will exert pressure on the condensed state that is called partial pressure whereas the chamber pressure or if it is in room that is called the atmospheric pressure or room temperature which is the p capital P okay this is called p star and we know the larger the temperature more molecule atoms or molecules will come so pressure will be building as a heat some so it is found I will prove this and that is what the proof I want to give p star is only a function of temperature for a species this I will prove by maths or physics whatever it is maths is always needed for physics p is defined as ambient hydrostatic pressure which is normally around other than the condensed state wherever the net pressure is is called the p according to Hertz principle or Hertz law as it is called the evaporation rate is proportional to difference of partial pressure and the vacuum pressure outside okay exactly at the difference of that in general in vacuum p is taken close to 0 it is not 0 but close to 0 compared to p star it is always very small so one can say it is proportional to partial pressure Hertz says that evaporation rate is proportional to partial pressure difference he said but I say okay p is normally close to 0 but we may not assume it we substitute that value and if mathematically it will get subtracted accordingly 6 decimal bar accuracy these days computer has 128 bits of accuracy so you can use it then there are 2 more terms we define Vg and Vc Vg is the volume of evaporation in a gas phase and Vc is the volume of evaporation in condensed phase condensed phase means either liquid or solid mostly solid so their volume will be very small because they are solid or liquid can come confined to area gas moves moves away so the volume of gas phase is always larger than condensed state if we define delta H e as the enthalpy change going from one state to other the correct spelling is written here Clausius Clapeyron these are 2 scientists and in those days scientists only used to become engineers the first they will do physics and then they will build a system so according to this equation using thermodynamics dp star by dt that is rate change of or not rate temperature dependent change of partial pressure is proportional to enthalpy divided by t times Vg minus Vc differential volume typically as I say Vg is much larger than Vc why I say you because Vc will condense state either solid or liquid very small volume gas large enough so Vg is are much higher you can keep Vc also if I give a value is subtractedly smaller typically constant of its propellant is unity so it is delta H e into t into Vg this is typically dp star by dt delta H e is enthalpy energy of formation of atoms okay or retaining that atomic structure or atomic bonding that is given for a given material for that a given temperature so for a given temperature delta H e is known so I know dp star by this is this but I also know this gas which is now I created though it is not accurate to say it is I it is ideal gas but we still believe it follows ideal gas laws we say PV is equal to RT ideal gas law so P star Vg is RT where R is universal gas constant is that okay PV is equal to RT 7 standard may but I have a baby way change never so dp star Vc equation for you modify Kiyato dp star by P star is delta H e RT square PV is equal to RT colidia and then reorganize the term so you get a term which is dp star by P star delta H e by RT square dt I take integral of that ln P star is minus delta H e by RT plus some constant which I take ln C1 constant can be in any form so I take ln C1 then I write P star is C1 e to the power minus delta H e by RT a be more than a P star is only a function of temperature flashes clapperon has proved that partial pressure is proportional e to the power minus so e to the power minus infinity is how much 0 e to the power 0 is sorry if t is infinite this is 0 so 1 so larger the partial pressure larger the temperature partial pressure is larger so I first find evaporation rate so I first find evaporation rate now we can see from here there are two fluxes involved there one of course is the one from the source the flux which is going up but from the substrate which is not sticking there is a pressure P outside P star like this P on the top return some flux return some flux so this evaporator evaporates vapor pressure down is essentially P dependent external pressure of vacuum dependent where this condensed pressure is dependent on the vapor pressure at the condensed state just about that how much is that now using this if a is the cross sectional area through which this flux come then there is a statement made by Hertz this is 1 upon a dne by dt where any is the number of atoms involved or atoms or molecules sometimes rate change of this this is the rate of number actually going with the time that is not area what does that mean flux this is essentially a flux which can be given by 1 upon 2 pi mk t to the power half P star minus P0 this is essentially called Hertz Knudsen equation this is called Hertz Knudsen equation but this is as I say is the rate represent the evaporation rate so evaporation rate R is 1 upon a dne by dt 1 upon 2 pi mk t to the power half P star minus P and P is how much 0 so it is very small and it can be used or not but just now I also said that some flux may return okay so we say whatever is not sticking may come back not necessarily because of pressure alone something which did not stick comes down okay the ratio of what is coming down to what is going is called alpha V which is related to sticking sticking coefficient how much sticking coefficient I want one everything should stick nothing should come so alpha V is typically maintained close to 1 this is where oxidation but this is nothing this is a vacuum this is ideal gas law there was no gas law use there except at the last flux okay there were 3 different fluxes except in gas phase we use gas was the other was solid there is no solid here there is only vacuum but there the partial pressure was inside solid and inside silicon oxide and this is no partial pressure in solids okay this is partial pressure in vacuum okay on the condensed state okay so R e V which is the vibration rate is 1 upon 2 pi mk t to the power half P star minus phi if we say alpha V is close to 1 then you can neglect alpha V otherwise I have to give a coefficient 0.9 0.87 what are number generally people in most cases we get alpha V close to 1 most cases if your system is well designed and you have a good how do I improve the sticking coefficient anyone if anything has to stick here what how can I improve sticking reduce then it will move about no I want stick increase so after the heated actually it will pick up it has an energy to bind you need energy to bind so you provide substrate heating is that clear so as you increase 100 degree temperature most of the material will stick is that clear to you so there is a alpha is adjustable value okay let us talk of 2 things have you written down this this is very as I said I am writing all this formula for what purpose I keep saying my process simulator has all those formulas built in then why I am learning because if the new process comes you should know how to actually model it okay one of you may do maybe one out of 20 124 main 123 may not do but at least the availability of one person I know how it is or for him only this is there are 2 things as I say in the pressure P 1 of course is the if you use something called Knudsen cell where we can restrict the pressure so it can be made 0 there is nothing from outside pressure coming in so I can reduce that word the other of course as I said just now earlier I want to improve mean free path so this pressure is very crucial smaller the pressure larger is the vacuum and larger is the vacuum lambda is so the at least how much lambda it should be whatever is the distance between source and the substrate at least 3 times that you should have mean free typically I may say the distance may be 30 centimeter and a mean free path of 100 centimeter is necessary for having one directional film growths film and depositions I repeat typical distance between substrate is a 1 feet or 30 centimeters at least 3 times that should be the mean free path sigma is the cross section of atoms per unit area P is the pressure T is the temperature at which you are doing this rest everything is known so I can always calculate lambda mean free path for a given temperature for a given pressure and for a species which cross section is known to 10 to power minus 4 per centimeter square or centimeter square whatever area atomic volume a square cross section you can substitute that number I will give in case I asked you to evaluate I never said the paper till last day so I cannot myself tell you what I will ask only last day I decide I will ask this so in that case often I cannot say what I can ask lambda is KT upon root 2 pi sigma square P where sigma is the capture actually cross section of the atoms which we are looking at pi R square pi R square R is the radius everyone I am not sure I am not read plumber's book but must be I mean all these are not novel or something this is not first time it was done in 18th century to 19th century so most of things are known to everyone like this I picked up from physics book something I picked from thermodynamics book now after I picked up many others are also picked up from the same source they never told that I have done everyone is doing same some of them are of course my own derivations but I in our times it was not very important to tell world that I had derived it I had derived it I derived it now it is before you derive you say I have done it that is the only difference so typically R e V is alpha V P star upon 2 pi M KT then there is to make alpha V 1 there is an if you have this is your place where you are hitting the material to be operated I have another small chamber round whose or vice or the is a same as what you write everywhere a word and now you can see if anything goes out or comes in there is a little chance of return attacking anything it is called Knudsen cell it is little costly so most operators do not have that but I just tell you that this is possible you know see if you are buying label or any small Edward system you want to buy in few lakhs so all these features are not given as great as that the Knudsen cells are used if you are doing some better vacuum system like MB has MB if you see MB if you see a molecular beam epitaxy system you have the Knudsen cell okay and because they are putting 3 crores and above then they will put you design of a chamber around the source it should be degasable it should not actually release gas if release it should be first cleaned up so there are many features in Knudsen cell it should have a area which is adjustable which I should fix it right now so Knudsen cell is a Knudsen cell okay but our aim is not a operation rate our aim is to get films thickness I want to take how much so I say the growth film growth rate is decided by operation rate divided by into mass by density that is the growth rate okay film growth rate rho is the density so if I substitute this REVP from here this can be written as upon 2 pi KT rho star by rho P star by rho if I adjust for a given species KT all these constants I know I for a normal row I assume that these are the metals where rho is roughly this this comes out for aluminum at least this 5.8 3 10 to power minus 3 A M by T to the power half P star where A is this area of Knudsen cell or if they are Knudsen cell whatever is the this area itself is the area so if I want thickness what should I do just multiply by the time of evaporation so that much film thickness you ask you know film thick per centimeter per second this is the deposition rate going on whatever time whatever thickness you want you decide time and what time it should do then shutter should come so there is a timer outside which can automatically do this you set the time for the thickness you want it has also connected that way I am not shown but there is a film near the substrate on the both side there is a monitoring this thickness monitors they also receive same films and they are mostly transducers okay they are capacity transducers so we actually monitor how much is the thickness of the film we are say tactile sensor actually okay so pressure change what I can figure out how much is the thickness okay you know you are you know tactile sensors it the tactile where do you use keyboard which has a tactile sensor pressure sensors so tactile sensors if him bhajan hoga utta pressure bada dea monitor hoga so I know how much I have first calibrated so shutter just comes tops as soon as shutter comes power shuts off okay so it is not that in our system you do not have anything like this so your hand on the shutter your watch this is what we do the reason is I am not interested in numbers of them okay so what is the actual evaporation process is it start from melting the evaporant substance as you increase the temperature the vapor pressure increases this is summary of what I said and evaporation of substance starts is that clear source of evaporation could be either a point source or maybe extended point source if you are extended small area so little more mass needed because now there are multiple points evaporation most cases we take it point source and solve mass we wish to find now deposition slightly what was our assumption all molecules are traveling straight okay but there are wafers all around so some flux will go at an angle they may not interact but they certainly will not only travel vertical they know at an angle also they will travel so if you have a film a substrate like this the film thickness here and film thickness here will not be same is that clear if this is your target some splux is coming some like this so there is larger correct thickness here but not necessarily same as this so some angle which it is sustaining has an effect on the thickness spherical reduces that but still it is a position the reason there is spherical because this unless the nuts and cells is designed that it actually matches with that are it has the same problem equivalent right now I am assuming only horizontal substrate and hitting it so is that clear to you why I am why I am interested I want on a vapor a number of a for same thickness all it can be if you put the source up down it will come down there is no problem but you just look at it if it is your evaporant in liquid what will happen to it gravity join a niche a no solution niche a jaya it is lay up or say I thought over a jaya you are all possibilities are true only think of all possibilities if you are solid source it will also fall but you can put a clips if you put a clip what material I should use is that clear so there are issues in this gravity so on the rucksack it does not bother us simplicity of system but it does not matter if it comes down or it comes if sigma is the solid angle over which source is emitting solid angle all directions is that clear there is a solid angle associated it is called omega the and raise the density of the apparent substance and these are the distances I nominated these are H vertical this is the distance at L where I want thickness and this is the R okay and let us say it is a standard angle to vertical as theta then since it is this all around so I will assume here and here it will be same here and here it will be same where theta is it will be same is that point spherical so everywhere same arm so all those position ending and R will have same thickness is that clear what I want I want to find thickness so roughly what I am doing is I assume a vertical parallel plate I measure it and put an angle to that is what I am I did not show you exactly in formula but the real is I will actually do as a vertically in one direction it come and projection of that along that axis cos theta of that this figure is 100% must be given in book so I am not saying this is because this this is needed how much thickness round how I thought it is projection it is an example yeah we have to project assuming it is vertical and at an angle what is the projection so that is the method mass people do so I thought I will show whether it is same theta is this theta and this theta that same yes they are only one Nina a poor a such a guy cone solid if vacuum is better better than minus 6 star mean free path is as good as 100 since typical distance between source and target is typically 30 please do not think it is a hammer a smitho pachar centimeter then you go there so chai then accordingly you push words so H is typically 30 centimeters we can assume flux transport reaching target with scattering at an angle at at low pressure return flux target it is it is very very small nothing P is not a pick then is that solid angle is how much now 2 pi surface and either because 2 pi 360 degree the solid angle is 2 pi is that okay the flux density reaching a vertical case is our operation by area to whatever 2 pi r square the growth rate is r upon 2 pi r square m by rho from geometric considerations what is geometry triangle h square plus l square is equal to r square and cos theta is h by r you want to test a minute h by r is cos theta theta is the angle from the vertical so cos the base is on the vertical base by hypotenuse is cos so h by r and geometry it is a since it is a right angle triangle h square plus l square is r square what I am asking is what is that I am looking into film growth rate at an angle theta our operation upon 2 pi r to h or a term should be on the angle please check it because I when I solid I do not check it may be sometimes I make me shifts do read the book and verify I am right and tell me also I am right why I say so at least I know you are read you read from the book and tell me I am right and if not that is good then I will modify actually I am writing it so much because I am actually day whatever lecture I prepare I just think I am writing there so I do not have to redo it I should not do this method I should only say give points and you should think but the way I am preparing myself I am just writing without saying anything so okay for your sake I gave you 3 lectures today and the earlier 3 or 4 whatever they are they will be available on moodle in next maybe 3 4 days okay but I keep telling you and my undergraduate students will verify that so growth rate any point l from the vertical point is r of m cos theta and this cos theta as I said with h by r r is h square l square under root so cos theta is h by r cos theta by r square h by r cube cos theta by r square is this sorry r cube no no no r square okay then the film growth rate is r operation m upon 2 pi rho h upon s square plus l square to the power 3 by 2 this r cube r i is 3 by 2 r square to the r r cos theta by r square is h by r cube r cube the current 3 by 2 term I now know how much is the film away from if h is the distance and l is the point where I am measuring thickness I know roughly how much is will be smaller or higher denominator is higher larger than the numerator obviously the maximum flux will go straight some will go on this that is why that planetaries are required and rotating systems are required so if everywhere gets every position finally will be seen equivalently on an average and therefore uniform film thickness last material will start even next time okay is that formula clear any point away the flux reaches is smaller okay so thinner thinner films and uniformity I will get if our planetary motion it rotates vapor rotates so everything is roughly seen uniform from where that was designed is this formula this was used to actually design the planetary system how much are we okay so what rate you should move so that this actually goes uniform so this is not just for the sake of it it was used for design there are typically there are two kinds of sources we use for this one is resistors heated even heated even we will see later but today only we will show you resistance usually the heating source is made of wire filaments not wires wire filaments of either tungsten tantalum molybden many others what are these metals called transition metals very unlikely for they will I mean their temperature of melt is 3000 plus okay now if I want to if I pass current through this tungsten or tantalum or molybden filaments they will heat because they are high square r it will heat okay the maximum temperature which most of them can give for a typical currents of the order of 200 amps to 300 amps is 1800 degree centigrade most metals which we use can evaporate at 1800 degrees typical depression rate using a resistive sources around 1 to 25 amstrongs per second this is I square r there we are there current I square r simple wattage the materials which can you if you do Jules it will be much better but you it is not Jules there is no second material on that okay this is not Jules Jules actually is done because of cold heating as we called okay some other time okay the materials which can be evaporate over this technique are aluminum 10 chromium antimony germanium indium gold silver magnesium calcium any species it can also deposit so-called compounds cadmium sulphide lead sulphide cadmium silicide where and even calcium fluoride where do these first three are used anyone solar cells solar cells then we can deposit in a CL KCL magnesium fluoride calcium fluoride calcium fluoride is very important material for self assembly nano structures for a pro calcium fluoride is a one of the biggest candidate for self assembly nano structures okay I repeat most things can be operated which has a melting temperature around 1800 can be evaporate anything higher than 1800 or 2000 it cannot be evaporated just the last slide for the day the typical filaments shapes are something like this there is a hairpin kind of this then you have a wire which you want to operate when it starts melting it will actually cover up all this wire because it will first liquefy the surface tension will increase and then it will start sometimes you have a wire bigger wire and larger flux you want you have a some kind of a coil and you put wires inside coil farmer things this is smaller than this it just fits inside if you have a you want something species powder then you can use boron nitrite crucible so it is not necessary that what shape you give me shape I will actually create the kind of filament it can generate the power currents are passing through this you need a current transformers which can give you typically 200 to 300 amps of currents thank you for the day