58223_en-US

liliconcom

Sliding Levels

Hello community - I am working with a split plot design with 1 whole plot factor and two split plot factors. The whole plot factor is a function of one part of the composition of the system, while the two subplot factors are two different functions of the entirety of the composition of the system.

 

Because the subplot factors are two functions of the composition, but of different properties relating to the composition, they are correlated.

 

I have found a lot about sliding levels in papers, but it doesn't seem that JMP has implemented it. Are there any good software solutions to this kind of problem?

 

Thanks

Re: Sliding Levels

There is a reasonable chance that I might be vastly over simplifying your experimental system. 

 

It sounds like your inputs are.

1. batch of water*.  Unless all the experiments are done with the same batch of dd di water. Then leave it out.

2. pH level (or maybe Molar equivalents of acid or base added)

3. not sure, if additional salt (ions) are added to achieve a specific ionic strength, but that would count.

 

-Molar equivalents of acid or base is a continuous varialble.

-Quantity of ions to adjust ionic strengeth is another variable.

-The solid added to the solution (mass if it dissolves, or surface area if it doesn't)

So we are looking at Three Factors, and I'd probably start with an RSM model for the DOE.

 (DOE, Custom Designs, when you get to adding model term, just click the RSM button)

Then the response is the equilibratiion curve. Which of course will need to be parameterized, and the parameters will become the responses of the experiment. (maybe a 4p fit?)

 

Over Simplified?

 

 

*after the Nature paper this week on water heterogeneity, I don't think I trust water anymore.

 

 

Re: Sliding Levels

Hello - Thank you for the reply. I am sorry, I did some other experimentation and forgot to check on the replies to this 'till now.

 

I don't know how much you know about this stuff so I will try to whittle it down, but I suffer from the common issue with scientists where I sometimes talk too technically. So please forgive me if I don't boil it down enough.

 

I am determining the change in acidity (pH change) of an electrolyte solution over time. My response is the final pH, and my factors are 1) initial pH, 2) Ionic strength (basically the concentration of ions in the solution), and 3) the chaotropic nature of the solution (chaotropic/kosmotropic ions disrupt/promote hydrogen bond interactions between water molecules).

 

Batches of water are each changed to different pH values (whole-plot), and then salts are added to adjust the Ionic strength and chaotropic nature of the solution (sub-plots). Then a solid is added to the solution that begins the pH changing process, and the pH change is monitered to its conclusion.

 

P.S. In case you want to see the relationship between factors, here they are.

 

Useful variables to know are:

a) m_i is the concentration of the ith type of ions in the solution, and

b) n_i is the number of the ith type of ions in the solution, where n_i = m_i*V (where V = the volume of the system)

c) Z_i is the charge of the ith type of ion.

d) dG_i is a parameter for the ith type of ion that is known from liturature.

 

The factors are obtained by

1. pH is measured experimentally, but
   1. m_H+ = 10^(-pH), where m_H+ is the concentration of Hydrogen ions and thus,
   2. m_i = m_H+ + all other m_i's
2. Ionic Strength = SUM((Z_i^2)*(m_i))
3. Kaotropic nature = SUM(-742*dG_i*n_i) [=Sum(-742*dG_i*m_i*V)]

Re: Sliding Levels

Sounds like you might need to add a factor and used disallowed combinations.

Can you describe what you are experimenting with?