 Good morning. Welcome to this lecture in our course on Chemical Engineering Principles of CVD Process. I thought we will use this lecture to just quickly summarize what we have covered in the course and if you have any questions you can ask them also today. So if you look at the material that we have covered, we started out by first discussing the fundamentals of chemical vapor deposition. What are its unique aspects compared to other methods for forming films on surfaces, the particular advantages and disadvantages of CVD process and we also looked at the various types of CVD films particularly the crystalline polycrystalline and amorphous films and we looked at the various process designs and reactor designs for making CVD films. Again focusing mostly on the temperature distribution, the cold wall versus hot wall and also the pressure, atmospheric pressure CVD, low pressure CVD and ultra high vacuum CVD or plasma CVD. Then we spent a few classes talking about the various properties of CVD films and how they are to be measured, both quantitative analysis as well as qualitative analysis and the middle portion of the course we primarily focused on the thermodynamics, chemical kinetics and transport phenomena associated with CVD reactors. So we looked at a method that will enable us to find the equilibrium composition of CVD products using a free energy minimization algorithm and then in terms of transport phenomena we primarily focused on the so called macroscopic model where we formulate conservation laws for mass momentum and energy and use constitutive relationships to provide closures to these conservation laws and again we focused on the mass transfer aspects since that is the most central transport phenomena in a CVD reactor and we described the thick diffusion process that delivers material to the substrate on which the film is being formed and we also looked at various adjustments that we have to make to the diffusional deposition rate to account for what we call the analogy breaking mechanism such as thermal diffusion, homogeneous reactions and heterogeneous reactions. In particular we derived formulae for the correction factors that you need to apply when such phoretic forces and other external factors are present. In the last part of the course we looked at some very specific examples of CVD processes and CVD reactors such as we looked at silicon CVD, CVD of coatings, CVD of oxides, hot wire or filament CVD to make metal films and finally nano materials as well. So overall that is the portion that we have covered as part of this course. Now for the exam I think as I mentioned before I really want you to have a good understanding of the breadth of the subject that we have covered and some knowledge in depth particularly about the chemical engineering aspects which is the thermodynamics, chemical kinetics both heterogeneous and homogeneous as well as transport phenomena. We had also spent some time talking about CVD processes that are not intentional but are actually undesirable such as burnout of tungsten filament lamps and also hot corrosion of turbine blades. So I do not think we have covered those in any of the quizzes yet. So that is possible that there may be one or two questions in your term exam on those two and in one of the recent lectures actually couple of lectures we talked about multi scale modeling which is a very unique aspect again of CVD reactors because there are at least four major length scales and corresponding time scales that are present in any conventional CVD reactor and it is important to understand the differences between these scales and also how they are interlinked together. In the last lecture we talked about the hierarchical approach to multi scale modeling versus the concurrent approach. So for a given CVD film the trick is how do you relate something that is so surface specific as the growth rate of a CVD film on a surface to something that is so macroscopic in nature because all you really have control over is how much reactant you are feeding in, what the flow rate is, what the temperature is and what the pressure is. So that is something you are trying to control at a reactor level but on that basis you are trying to make a very very thin film or structure on a substrate. So you are trying to influence what happens at a nano scale with parameters that you are only able to control at a much larger length scale. So that is really the challenge in a CVD reactor. So unless you have a very precise understanding of how the phenomena are linked from the various length scales and time scales that are present in your system you can never achieve the kind of control over the nature of the deposit that you are forming on the substrate. So another question that I think is very relevant to ask in a term exam would be something around the multi scale modeling of CVD reactors and how you would apply that in order to make this linkage from the controllable parameters in your reactor to the functional parameters that you are trying to achieve with the reactor. And you know at the largest length scale we have the macroscopic model or the conservation model which is where we spent quite a bit of time on. So here again I think I would expect you to be able to discuss essentially from first principles how you would form a mass transfer model or film formation happening in a CVD reactor and identify the appropriate correction factors to apply for various reactor operating conditions. For example if you have a cold wall reactor I would expect that you should be aware that A in a cold wall reactor natural convection is going to be dominant because the temperature difference between the hot substrate and the cold wall is going to drive natural convection phenomena. And secondly I would expect you to be aware that the temperature gradient is again going to require you to include thermal diffusion as an influence in your deposition rate calculations. You have to be able to calculate the correction factor corresponding to thermal diffusion and apply it to your result. Similarly if I tell you that it is a hot wall reactor then I expect you to appreciate the fact that when you have a hot wall reactor the probability of heterogeneous nucleation is much higher and so the probability that some material will condense in the gas phase and potentially be transported as aerosols to the surface is also higher. So again I want you to demonstrate you know that level of understanding when does, when do homogenous kinetics become important, when do heterogeneous kinetics become important. And again central to all this is a figure that we have drawn several times which relates the logarithm of the deposition rate to the inverse of the surface temperature. You really have to be able to identify the various regions that are present in that graph and ascribe you know reason for the behavior that you see. Why is it that at very high temperatures you actually start to see a drop in the deposition rate. Why is it that at low temperatures you start seeing more of an exponential behavior with respect to temperature. In other words when does surface kinetics become important, when does gas phase kinetics become important and when is it that diffusional phenomena dominate over either. And how do you express it? You have to be able to express the relative importance of the various factors affecting CVD rates in terms of non-dimensional parameters. So you have to be able to define the Damkohler numbers, Nusselt number, Stanton number, Pecklay number. So if you look through your notes these are dimensionless quantities we have defined which essentially represent the ratios of characteristic times. So it could be the characteristic time of diffusion versus chemical reaction or it could be the characteristic time for homogeneous reaction versus heterogeneous reaction. It is important that you should be able to express the relative importance of the various terms influencing CVD rates on the basis of the relevant dimensionless quantities. Again that is a contribution a chemical engineer makes you know it is anybody can do experiments in a CVD reactor and look at the data and react to it. But in order to be able to interpolate those results or extrapolate those results you have to be able to define the so called dimensionless parameters which enable you to do this kind of scaling and extend the region of validity of your experimental data. Otherwise you can never be sure that you know whatever data you have collected for a certain set of conditions if you develop a model based on those experimental data now those are called empirical models. The problem with empirical models is as soon as you go outside the range in which the data were obtained you can never be sure that the same model is valid. And that is where the non-dimensional analysis becomes very important because we can then take the results that we get in one system and be able to apply it for multiple systems. So that is another value add that chemical engineers can bring to the study of chemical vapour deposition. So I think you know the point I made in one of the very early lectures is that a CVD reactor is like a microcosm of the chemical process industry you know everything is happening inside a CVD reactor everything that you have studied in chemical engineering. And I hope that as we come to the end of the course you have a good appreciation for that you should be able to assess how you know obviously the 3D transport phenomena that you have studied have relevance in the study of CVD systems. How chemical reactions and equilibrium thermodynamics have a relevance in the study of CVD systems and also the control methodologies which is another important subject that chemical engineers learn also has application in CVD reactors because the most important thing in a CVD reactor is to control the product which is the film. And in order to be able to do that you need to be able to control all the different parameters that affect the operation of the CVD system. So control methodologies are extremely important also in the investigation and optimization of CVD systems. So I hope that you have kind of been able to extract all this from both the lecture notes as well as the papers that I have been circulating. If you have any questions go ahead, is there anything that has been bothering you from day 1 about CVD or anything that has come up in the last couple of days or is there anything that you think should have been covered which we did not cover in class or something that we covered in class which we should not have covered in class. Any questions, feedback, anything is welcome. In nanomaterials, we have studied very small part, we have gone through little bit more. How to make that carbon powder nanomaterial or any other, how this... Yeah, by the way I have sent some papers this morning. So I have sent a bunch of papers on basically synthesis of nanomaterials. I do not know that there is lot to be gained by teaching that in class, I mean there is nothing very fundamental about it. And so I have sent you some papers, take a look at them and if you have any questions you can always get back to me. See, synthesis of nanomaterials, I do not really believe that CVD is the best method for doing it. It is one of the methods and it is something you use when you are not able to use other methods but I do not think it is the optimal use for CVD. I mean CVD is really good when you are trying to make coatings over surfaces that have complex contours and stuff like that and also very, very fine clearances. So it is well suited for microelectronics but when you get into the nano range, I do not think CVD works as well as some of the molecular techniques that are now being used. But yeah, I mean it is certainly something interesting to look at. Chemical sensors actually basically used for a place amount of gas addiction, nanomaterials are being coated on some surfaces then those materials are being used in actual purpose. In that case, this CVD I think is much more important. Yeah, certainly. Anything else? Maybe I will ask you, what did you think was the most interesting part of the course? Application of various conservation equations which we learnt separately, the combination of all those in CVD. You also have a course in transport phenomena, right? Yes. Don't they try to do that in that course? I thought that is where you try to combine the various. They did try to do that but I think maybe, maybe practically it was scenario where we had a, like we have a CVD system and just like you mentioned, like if it is a diffusion dominant or thermal diffusion dominant, maybe practical scenarios in that case. Okay. Yeah, one thing we did not really do a lot of is solving numerical problems, maybe that is something that can be done as well. What was the least interesting part of the course? This is where it got very theoretical, the spectroscopic microscopy and all that. So you would want to see some practical demonstrations that make it more interesting because it is important to understand how CVD films are measured, whether we like it or not, it is something that we have to know. I mean ellipso-metry accounts for probably 90% of CVD film thickness measurements. So if you are only going to learn one technique, ellipso-metry would be the right one to learn. Any, you have any comments or, okay fine. So let us break a little early today and if there are any numerical problems, I think they will be designed such that you do not need a calculator but just bring one just in case. Okay. Good. All right.