Virtually any of transition metal could. Cu, Fe , Carbonaceous preg-robbing ores or any other transition metals can consume your free cyanide.

Zn is very common to complex the cyanide as well as copper to cause such titration errors.

This is a clear solution sample I was talking about, so theoretically after pH raised from 9 to above 12, free cyanide concentration should increase, generally it does.
But just from time to time we found odd that it actually drops after raising pH, and by more than 30ppm? The other day, it went from 85ppm, dropped to 50ppm when I titrated after pH raised above 12. This solution sample would have Cu, Zn, Fe, Au and Ca etc. ions that leached out during gold leach process.

My questions are what condition/environment triggers this to happen and what to watch for, how do you explain it?

Copper and iron?

If the solution is free of metals then it is possible that some of the free cyanide ions react with sulphur in rhodanine to form thiocyanates.

I agree with you and will suggest that you compare your cyanide titration with one using Silver nitrate and Potassium Iodide - follow that up titrating for alkalinity using Oxalic acid and phenolptaline- there is always the chance of over titration and has to be kept in mind in all titration methods.

Just a suggestion! Look up the CN "complexation" equilibrium stability constants, and the Free Cn action in hydrolysis. At the level of ppm, concentrations, these reactions are quite significant. The problem seems (again at those concentrations) to do with hydrolysis, notwithstanding the complex formations and high pH would essentially drive it towards free CN. Have you had a chance to analyze the base metal concentrations? Although you suggest that there should not be any, a quick AA or ICP scan of the suspected analytic species would be informative and rule out this possibility. After which address once again the hydrolysis aspect, I will follow the thread.

Here is some analysis I got out of our historic data.

Cu CN_WAD Ca Co ECond Fe Mg Ni TDS_180C Zn
iMET1WCICP iCNW1WAAA iMET1WCICP iMET1WCICP iEC1WZSE iMET1WCICP iMET1WCICPiMET1WCICP iSOL1WDGR iMET1WCICP
0.002 0.01 0.1 0.005 0.2 0.005 0.1 0.01 10 0.005
mg/L mg/L mg/L mg/L mS/m mg/L mg/L mg/L mg/L mg/L
2490 0.54 7490 16 239 2.5 54000 0.33
2860 0.54 8050 16 125 2.2 58000 0.36
2660 0.55 7560 12 182 1.9 54000 0.53
2600 0.53 7380 12 165 1.9 53000 0.35

they seemed to be very consistent, are there any other elements that could contribute the odd that we should have but haven't analysed?

I find I cannot read/follow the trace metal concentrations from the list above. Is it possible to list these as:

Cu ppm
Zn ppm
Fe
Mn etc.

However if I am to interpret the data, is it fair to say that Cu is at a concentration of between 2400 and 2600 ppm (mg/L)? If so does this withstand the equilibrium constant for complex formation for Cu.

I do not have any tables available with me, but if the data from the various analytic species are plugged into the formula for the EQ Constant for the complex ion formation, would this be a possible area where the CN ion concentration is disturbed by the common ion effect? If so then the arguments and suggestions provided earlier in the thread seem justified. I also would then suggest that CN determination be performed in duplicate by two different methods. Titration methods have a tendency to be prone to human error. If you have access to auto analysis by a parallel method using say colorimetry (preferably by auto analyzer) then the two results can be compared. The question that also needs to be answered is: How significant is this phenomenon: does it have any economic implications? Does it have interferences on other analyses or products in the process scheme? 

Have you tried doing total cyanide after distillation on each solution? If the cyanide is complexed in the high pH solution but not the low, then the results should be the same after distillation. If there is a difference in the cyanide concentrations that suggests that the cyanide has been converted to some other species. Is it possible that some of the cyanide is just consumed in the process?

PKa of HCN is c. 9.3. Most likely you are getting a false positive reading owing to sulfide present in the solution. How are you making up the rhodamine and what is the tiger of your silver nitrate? Suggest toy run metals by ICP.

That is in sync with what I have suggested above. With tailing solutions, high pH requesting an ICP scan should include advising the lab for possible interferences with high Ca and Mg.

If you have done elemental analysis of ore it would be predictable to suggest competitive reactions of metals (Fe, Zn, Cu, and Au with their % age composition) with cyanide. Free cyanide is usually estimated with silver nitrate solution.

I believe that you already have the analytic species in solution when determining interferences at the point of analysing CN concentrations. Once again an ICP scan for the suspect analyses can provide a first set of possible interferences. I also think that if there is the complexity of elemental content, a more specific CN analysis method is required.

Agreed! ICP scan will definitely provide a clear picture of metallic contents. Yes, I mean it is required to get AAS or ICP analysis of an ore. After having an idea about metallic percentage, it will be convenient to know about free cyanide. Stoichiometric data can be helpful in this connection.