Kinetics of chemical processes

[#1]

Hallo,

on one side there is an impressing number of possibilies of JMP to plan complex chemical experiments for the study of a reaction and to evaluate the results by nonlinear, multivariate and linear regressions. On the other side, as a chemist, I wonder how or if chemical reactions with an exponential dependence of the temperature according to Arrhenius k = k0 * exp(-EA/ (RT)) or complex reaction rate laws such as dr/dt =  k1(T)*[A])/(1+ k1(T)*[A]+k2(T)*[B])  with concentrations of A and B on the catalyst can be evaluated kinetically by such possibilities. Can I get not only the optimised conditions but also the kinetic parameters? Only for thermal reactions or also for electrochemical reactions? 

Only with a kinetic model the precise description of a reaction is possible. Do I need another software package to obtain kinetic data?

Or can I just perform a definitive scanning design and access all desired data ?

Sincerely

Thomas Willms

[#2]

The Definitive Screening Design is intended for cases where you have many factors and little a prior knowledge of the relationship between them and the response. A linear model is hypothesized for the optimal design.

You have a non-linear theoretical model with physical parameters. I recommend you study the chapter about the Nonlinear Design platform.

[#3]

I'm not sure exactly what you are asking, so my comments may not be useful.  Here are my thoughts:

1. If you are asking, can you experiment on the variables of theoretical equations to determine if those theoretical equations "work" in your specific application, the answer is yes.  It doesn't matter what type of mechanism is at work, as long as you can quantify/measure the results, you can experiment on the input variables.

2. There are a number of elements you will need consider (e.g., How will you handle noise, what levels to manipulate the factors at, how many factors will you manipulate, resolution, restrictions on randomization, etc.) to determine the appropriate design strategy.  I always recommend you design multiple options and compare each design for potential knowledge gained vs. resource requirements.  JMP is great at creating the designs quickly.

3. Analysis is always Practical (from the eyes of the chemist) then graphical and lastly quantitative.  JMP can certainly help with the analysis.

[#4]

To add just a bit to the thoughts of @statman and @Mark_Bailey if you are evaluating your responses over time during a reaction or event, perhaps JMP Pro's capabilities in the Functional Data Explorer platform can help with the analysis. Here's a link to the documentation: Functional Data Explorer 

[#5]

@P_Bartell, I thought about using the Functional Data Explorer, too, but rejected it. It is not based on a theoretical model. It uses interpolators or specialized functions for spectral data. This case demands fitting a theoretical model to support the chemistry interpretation through the physical model parameters.

[#6]

Understood...I was more thinking of just the compact visualizations that come with FDE than the actual modeling work. I agree with the idea of a theoretical model with non linear parameters is going to best be handled in the non linear platform. I was always a proponent of throwing the kitchen sink of JMP's many platforms and workflows to see what one can 'see'.