Kps of silver halides

[mghis]

Hi all.

Can anyone post a table (or a graph) of solubility product of AgBr, AgI and AgCl at different temperatures? Kps at 25°C are given in most handbooks, but I have not found values at higher temperatures.

(I think I could derive them using van't Hoff equation but then again I should have to know how ΔH° varies with temperature )


Thanks!

 

[Photo Engineer]

I can give you some information right now, but not what you want.

I hope this helps a bit.

PE

 

[mghis]

Thank you! (I think you posted KBr curve twice, and no KCl curve)

I was trying to reproduce the calculations of the "example" section of a patent I found (5,670,282). The patent gives only the results.

 

[Photo Engineer]

I can calulate pKSP, but need some time. I can also use equations to get what you want, but we usually use vAg or pAg. Let me know what you need.

PE

 

[mghis]

Well, if you could draw up a table (or graph) of pKsp vs. Temperature I think it will be useful for many readers.

For instance I was trying to reproduce calculations of Example B of US5670282 patent (link here). In this example, KCl was added to a kettle to get vAg = 150 mV at 35C and then 0.01 moles of AgNO3 and of KCl were added. Then it should precipitate 0.01 mole of AgCl. After 10 seconds, six liters of solution containing 0.0099 moles of KI were added. vAg stabilized to 150 mV. Using the method described in the patent we should be able to predict that solid phase composition is 99.2% AgI and 0.8% AgCl.

I think that using Nernst I could get aAg+ (silver ion activity).

E = E° + RT/F ln(aAg+)
ln(aAg+) = (E - E°) F/RT
aAg+ = exp [ (E - E°) F/RT ]

If I got it right, it should be (for vAg = 150 mV) aAg+ = 4,05 10-8 .

We also need to know the Cl- and I- concentrations. I thought we could get them using Ksp = aAg+ aX-. But Ksp is needed. And I am not sure this is the right method...

EDIT: the quantity of KI added is written as "9.9 mmoles". I formerly interpreted it as 9.0 milli-moles. Now I begin to think it could be a typo for 9.9 moles...

 

[Photo Engineer]

I'm familiar with the patent and I know both inventors personally. I reduced that work to usable software for making and scaling. We did not use PKSPs to do our daily work, so I would be interested in knowing what you are trying to do.

In your case, Cl vs I, just assume all of the Iodide is precipitated, as there are at least 2 orders of magnitude between Cl and I solubility at temps that you would use. This is close enough for our work.

PE

 

[mghis]

I thought that I needed to know Ksp because they are in equations 6, 7, 8 and 9. But if you did not use them, there must be something that I'm missing.

So all of the iodide is precipitated. We added 9.9 mmol of KI so we should have 9.9 mmol of AgI. I think we don't know Cl- concentration: we know that at the beginning it was 0.0125 M. After the first run of AgNO3 + KCl it should remain roughly the same if the additions are equimolar. But then we added 6 liters of KI solution. We didn't know the original solution volume so we don't know how much Cl- gets diluted.

 

[Photo Engineer]

Well, you should keep track of all volumes, but in this case the Cl amount should be Ag moles = I moles + Cl moles pptd. Therefore the remainder is all Cl. That is a close approximation. If it were Bromide though, all bets are off. The PKSPs are close enough and there are solvent effects, so this is all upset. A Chloro/Iodide is a very special case though and used for a very special purpose. They are often hard to make (control) and hard to sensitize.

If we needed any value, we had a "toolkit" that gave us those values. It calculated everything for us and put it on-screen or printed it out. The trouble is, it will not run in anything above Windows ME due to the age of the code. Otherwise, I could probably run some simulations for you.

Best of luck.

PE

 

[mghis]

Thank you for your help, PE!

It's a pity, though, that you cannot run anymore your program. If you ever want to try, there are a few programs which can perfectly simulate old operating systems, like DOS and MS-Windows (all versions of it), and UNIXen. Two of such simulators, freely available, are VirtualBox and DosBox. DosBox will emulate DOS and Windows versions that run "on top" of DOS, like Win 3.11 and Win 95. VirtualBox will emulate every operating system. From my experience they work really fine. If you try to set them up and encounter any problem I'll be glad to help if I can.

 

[Nodda Duma]

I would love to try to get that s/w up and running (and then, of course, figure out how to use it)

Ron, maybe I can find a cheap Windows ME laptop to send your way. I’m pretty good at scrounging up stuff like that

 

[Photo Engineer]

I have an ME laptop thanks Jason. I also have VMware player for my machine. I can run it, but there are restraints in memory and also in transferring results to this machine. I use "Wormhole". I have a whole nest of problems. Windows 10 will not run ME so far in emulation mode, but I am working on it.

As for the results, well, it is not really useful without an understanding of Chemistry IMHO. So, who is up for that. I have equations that I published in the book. Look at them. Glenn's work is correct and it works. Try it. You can probably write something useful before I can emulate it.

PE

 

[bdial]

Creating a VM and installing an OS on it with VMware player is tricky but possible. I've done it, but it's been a while.

Can the software run on Windows 2000?

 

[mghis]

Of course I believe that Glenn Judd's work is correct. There are just a few things in the patent that in my opinion are dubious and might be typos or omissions. For instance, in example A, activity magnitude order appears to be unreasonable: it says 10^11 molar! (In fact this is repeated consistently and I don't know what the inventors really meant, this is why I tried to make sense of example B)

Regarding PE's program, I don't think it's necessary to set up a new simulation for example B. I think that simulations of more conventional emulsion types that you have already run will be even more interesting for us all. Then we could try to roughly repeat calculations by hand and see if we can get anywhere near the real results.

For file transfer between "inner" emulated OS and real machine, I usually set up an FTP server inside the virtual machine. Then I connect from my machine to the virtual machine with an FTP client. But I imagine that Wormhole will do just as fine.

 

[Photo Engineer]

I just use drag and drop. It works fine from the emulated machine to the desktop.

However, I think I may have some examples here, and I'll try to find them.

PE

 

[Photo Engineer]

I found an example and information if anyone is interested..


PE

 

[mghis]

I'm very interested!

(Moreover, I'm looking forward to search in my university library data about silver halides Ksp.)

 

[Nodda Duma]

Always interested, Ron.

 

[Photo Engineer]

The sample I have is a BMP capture of a virtual screen. I am sorry for the color. It was a DOS image and that should explain a lot. The colors were selected by the customers. Ahh, emulsion makers.

This is a single halide example. The multi-halide example was a complex calculus equation that ran as a function of dF/dT or flow rate vs time. It was not static. It ran as time elapsed and you saw the values changed. You could pause it and resume.

PE

 

[mghis]

Well, that's really neat indeed! Now I'm going to crunch those numbers for a while!

If you happen to find other data like this, please post them if you can. We use to say that appetite grows with eating.

BTW, if this is a DOS-only program, you might want to try DOSBOX. You don't even need DOS installation disks: you put your tool in a directory in your machine, like C:\KEDS, you fire up dosbox and type in "MOUNT C C:\KEDS", press enter and then "C:". Then, (for every DOS program I have tried so far), you are ready to go! It runs on your machine's disk, so you don't need special provisions for file transfer.

 

[Photo Engineer]

Yes, I have several DOS boxes. However, when I do work, I am trying to do so much more as noted. dVag/dt and etc.....

PE

 

[mghis]

I finally found something about pKsp! From Owen and Brinkley, J. Am. Chem.Soc. 60, 2233 (1938).

According to their work:
pKsp(AgCl) = 5905.8/T - 22.8538 + 0.057854 T - 5.0131 10^-5 T^2
pKsp(AgBr) = 5702.2/T - 11.1822 + 0.014610T
pKsp(AgI) = 7084.0/T - 12.0389 + 0.014610T

Unfortunately they ran experiments from 5°C to 45°C, and emulsion precipitation temperature can get higher. Well, extrapolation is better than nothing...

This work seems to be in decent agreement with KEDS screenshot. It says that at 40°C pKsp(AgBr) = 11.6105, while Owen and Brinkley predict 11.6086.

 

[Photo Engineer]

There was a long thread with lots of this information in it, but Sean deleted it years ago because I argued with another member over these values. All I wish to say now is that the Kodak values worked.

PE

 

[mghis]

I did not see that thread. It is unfortunate that is was deleted. It's very rare to come across this kind of informations. I do believe that Kodak values work. Kodak has been making silver halides for 130 years by now! It's most likely that the values of the article I found are less accurate, but they'll have to suffice, since I don't have EK values.

Regarding KEDS screenshot, these are my thoughts:
1) Ksp = a(Ag+) * a(Br-) therefore pKsp = pAg + pX. Plugging in numbers, it works fine
2) vAg value follows from Nernst equation and pAg. Also, I replicated KEDS result by hand, but...
3) I have no idea of how junction potential was calculated. I think it depends on electrolyte used in salt bridge and on its concentration. I'm not sure. My textbook only spends few words on it. Anyway it's a minor correction.
4) I tried to calculate activity coefficient using Debye-Huckel limiting law and I got a slightly wrong value (I got 0.6903 instead of 0.77163). I imagine that KEDS uses some form of Debye Huckel extended equation. It would be interesting if you had more information on this.
5) I tried to calculate pX from halide concentration and given activity coefficient and got a slightly different result. I cannot explain this and confuses me.
pBr = -log a(Br) = -log ( bromide conc. * activity coefficient) = -log (0.1 * 0.77163) = 1.11259
Instead KEDS prints out 1.1107. I can't understand why.

 

[Photo Engineer]

Juncpot = 2.5 * rtf * (z1 / (z2 - z0)) * (log z2 / z0) / 2.302585
rtf = ,198415 * tk
tpot = 603 - (,112 * (t - 25)) + (.0008 * (t - 25) * (t - 25))
brpksp = -12.1258 + 5868,64 . (T + 273,15) + ,01593 * (T + 273,15)

Does this help?

PE

 

[mghis]

Thank you!!
I think I might need some help with variable notation. "brksp" is pKsp of AgBr. "Juncpot" is junction potential. z0, z1, z2, I don't know (but since it is used for junction potential, might be concentration of reference cell, salt bridge and reaction kettle?). "tk" I guess is Kelvin temperature? "tpot", also I don't know. Plugging in T = 40°C I get tpot = 601.6. Seems to be a potential in mV. Maybe reference electrode potential?