Overclock.net › Forums › Cooling › Water Cooling › copper + brass + steel + solder loop — glycol mandatory? Also IBM and benzotriazole
New Posts  All Forums:Forum Nav:

copper + brass + steel + solder loop — glycol mandatory? Also IBM and benzotriazole

post #1 of 3
Thread Starter 
Is it mandatory to run a propylene glycol coolant for something like the EK kits? (I'm not interested in ethylene because of the toxicity.)

They have:

steel (the jet plate)

I've been reading so many older comments about people with "all copper" loops. How does that happen? Does someone make copper jet plates? What about copper fittings? Are there radiators with no solder that comes into contact with the water? I guess this might have been the case before jet plates and micro channels. But I have yet to see a copper fitting anywhere.

EK shows scary corrosion on its website for distilled + coil loops. I've looked into various anti-corrosion additives and the most promising might be benzotriazole. IBM is using it – and without glycol — which seems to be a pretty big vote of confidence:
Originally Posted by IBM 
Deionized water with benzotriazole solution installation and maintenance
IBM is responsible for supplying and maintaining the internal frame secondary loop side water. This protects the system processor books and distribution plumbing from damage that could result from the use of contaminated water. IBM supplies a water solution that is mixed with benzotriazole (BTA), a corrosion inhibitor, for the internal secondary cooling loop of the frame when it is installed and when any repairs are performed that require water to be added. IBM uses certified suppliers for the water solution that satisfy all pertinent environmental control requirements.

Deionized water with benzotriazole disposal
The customer must dispose of the water solution in accordance with applicable laws and regulations and product characteristics at the time of disposal.

Internal frame water solutions
IBM will supply the system-side water.

Water and benzotriazole solution
Benzotriazole (BTA) is mixed with the deionized (DI) water to a concentration of 1000 parts per million by weight.

Deionized water

The deionized water used in IBM water cooling systems conforms to type II, grade A specifications in ASTM D1193-06 entitled, standard specifications for reagent water. Specifications are as follows:
• Electrical resistivity at 25 °C > 0.5 MΩ•cm
• Total organic carbon < 50 µg/L
• Sodium < 5 µg/L
• Chloride < 5 µg/L
• Total Silica < 3 µg/L
• Total organic carbon < 50 µg/L
• Heterotrophic bacteria count (HBC) less than 10 colony forming unit (cfu) / 1000 mL as measured per ASTM F1094 or IBM approved equivalent


Benzotriazole (BTA) is purchased from Sigma-Aldrich or an alternative IBM approved supplier, and is defined as follows:
• Product Name: Benzotriazole, 99%
• Product Number: B11400
• Brand: Aldrich Chemical
• Substance Name: 1H-Benzotriazole
• Chemical abstracts service number: 95-14-7
• Formula: C6H5N3
• Molecular weight: 119.12

Apparently benzotriazole is difficult to get to dissolve, though.
post #2 of 3
Thread Starter 
Originally Posted by superstition 
Is it mandatory to run a glycol coolant for something like the EK kits?

Apparently EK doesn't think so since their pastel coolant (Mayhem's rebranded, reportedly) is 40-60% glycerol rather than a glycol.

Is this glycerol actually for corrosion prevention or just to keep the particles in suspension (aesthetics)?
post #3 of 3
Thread Starter 
I asked about the practice of using iodine as a biocide in loops but have discovered that it is also potentially useful as a corrosion inhibitor (in the form of potassium iodide) when combined with benzotriazole:
Originally Posted by source 1 
Potassium iodide (KI) has been used in combination with copper to form polymers that could be used in laser protection systems (Li et al, 2006: 1415). Potassium iodide has been tested as a means to improve the inhibitive efficiency of BTA in sulphuric acid solutions because of the synergistic effect observed between them. In general halide ions can improve the adsorption of organic inhibitors on the metal substrate resulting in increased inhibitive efficiency of the compound. Therefore the presence of potassium iodide enhances the adsorption of BTA on copper (Schweinsberg et al, 1997: 161).

Wu (1993: 215-217) investigated the synergistic effect between BTA and KI for copper inhibition. The adsorption of iodide on the copper substrate was found to occur prior to the formation of the protective film. The addition of KI not only enhanced the adsorption of BTA onto copper due to an electrostatic force between adsorbed iodide anions and BTA cations but also resulted in the formation of a new copper iodide-BTA (CuIBTA) complex with a molar ratio of iodide to BTA of 1:2. The film formed was thicker than that o f BTA alone. There was no information as to whether this complex is polymeric or not in its structure. One important issue is that the coverage of the surface was not uniform and more importantly the efficiency was affected by the formation of the film.

Wu et al (1993) suggest that since BTA is found in a protonated cationic form in solution (BTAH2+) the presence of iodide on the copper surface could improve its adsorption therefore its efficiency. Iodide could possibly form a dipole with copper orientated with the negative end in the direction of the solution. Possible explanations of the synergistic effect observed between BTA and KI could be that it is due to an electrostatic attractive force between the adsorbed on the metal iodide and BTA that improved its adsorption. Alternatively it could be the formation of a covalent bond between a noniodized BTA molecule and the adsorbed iodide, or a positive shifting of the zero charge potential that leads to negative charge of the metal surface resulting in improved adsorption of BTA on the metal

(Wu et al, 1993: 2791, Schweinsberg et al, 1997: 161). According the Wu et al (1993: 2799) the addition of KI in the BTA solution resulted in the formation of a thicker, better polymerized film that incorporated iodide, Cu(IBTA), affecting the anodic reaction:

pCu + ml- + nBTAH -» Cup(ImBTAn) + 2nH+ + pe’ (29) pCu + ml- + nBTAH -> Cup(lmBTAn) + nPT + pe (30) Where p = m + n.

The reaction of the protonated BTAH2+ (equation 29) is most probably favoured since the amount of BTA cation is predominant in solution by four times and iodide forms a dipole with copper. Iodide was found to react with BTA only in the presence of copper forming a cuprous iodide-BTA polymer (Wu et al, 1993: 2791-2800).

KI on its own resulted in accelerated corrosion since the formation of a Cul film was proved to be unstable (Wu et al, 1993: 2793).
Originally Posted by source 2 
Potassium Iodide acts synergistically with BTA. Different theories on its mode of action, formation of Cu(IBTA)

BTA and KI

Wu et al (1993) suggested that the presence of iodide ions in the BTA solution improves the adsorption of the inhibitor on copper. The presence of KI in the BTA solution results in the formation of a thicker polymeric Cu(IBTA) film affecting the anodic reaction (Wu et al, 1997).

The combination of BTA with KI was tested taking three factors into account: BTA concentration, KI concentration and solvent (deionised water and ethanol). The experiments showed that the results and the possible synergistic effect between BTA and KI depend on their concentration and the solvent used. Consequently, the treatment of copper with this combination can lead to im proved inhibition or increase of the corrosion rate. The mixture of 0.1M BTA with 0.01M KI in deionised water increased the efficiency of BTA, whilst in other cases the mixture accelerated corrosion. When a higher concentration of BTA was used, the presence of KI did not cause any significant effect in the process.
The bolded part is concerning since it seems like a delicate balance of the two chemicals may be needed.
Originally Posted by source 2 
From the images it is evident that some inhibitors follow the crystal formation of the corroded copper surface, forming a thin film (e.g. AMT, BTA, ETH), whilst in others (BZA, KEX, KI, PMT) the copper crystals are not distinctive anymore, as the surface is covered with a thick porous non-uniform layer.
Note that the presented scans are all in ethanol solution rather than deionized water which may impact the film.

Edited by superstition222 - 10/3/16 at 10:02pm
New Posts  All Forums:Forum Nav:
  Return Home
  Back to Forum: Water Cooling
Overclock.net › Forums › Cooling › Water Cooling › copper + brass + steel + solder loop — glycol mandatory? Also IBM and benzotriazole