Further things which may be of interest to members

[Deano]

DIBK ALIQUAT 336 SOLVENT EXTRACTION METHOD FOR GOLD Atomic Absorption Spectroscopy
Digest sample in a 500 – 600 ml beaker containing around 250ml of 1 part nitric acid to 4 parts hydrochloric acid.
The aqua regia must be freshly made, do not bottle for later use.
The beaker is heated to a simmer on a hot plate in a fume cupboard, orange brown fumes will evolve.
Continue the heating until the fumes change colour to white, give it a few more minutes and remove from hot plate to cool.
When cool the solution is filtered, usually in a Whatman No1 paper and the paper is rinsed with water from a wash bottle to get all traces of gold into the filtrate.
After filtering has finished the filtrate is made up to usually 250ml with water and poured into a clean beaker to evenly distribute the gold values in the solution.

In a clean 50ml glass extraction tube add 50ml of the final filtrate and then add 5 ml of DIBK Aliquat 336 (1%) ( straight hypochlorite solution is ok, but Aqua Regia must have Nitrates expelled as above). This is a 10:1 concentration ratio.
Shake thoroughly 100 times. Allow to settle. If there are drops of solvent adhering to the walls of the extrtaction tube the tube is sharply but gently finger flicked to dislodge those drops and have them join the organic phase.
If any haze is seen in the solvent section, or there is concern of excessive iron in the sample, add drop by drop some concentrated hydrochloric acid in a very slow and controlled manner dropwise with a 2 ml or similar dispenser. The now clear DIBK solution at the top of the extraction tube is ready to scan directly on AA.
Special note: The standards used on DIBK extraction must be made of DIBK Aliquat 1%, eg. The 0 ppm reference is straight DIBK Aliquat 1%, and the 1ppm and 5ppm need to be made up from straight DIBK Aliquat 1% solution and addition of Gold chloride standard.
Flame adjustment is only needed if maximum sensitivity is required , the flame can be of lesser intensity for DIBK than the flame used for direct aspiration of aqueous liquors.
Usually the small improvement in sensitivity when using DIBK is not worth the fuss of flame adjustment and the flame is set so that aqueous samples may be direct aspirated with out flame out and this setting is used for all liquids.
NOTE THAT THE DIBK STANDARD READING WILL DIFFER FROM A GOLD CHLORIDE OR CYANIDE READING DUE TO THE DIFFERENCE IN UPTAKE RATES FOR THE SAME STANDARDS.
**This method removes all Iron and interfering impurities from the leach to negligent levels even in saturated Iron solutions. DIBK solvent cannot be used on it’s own but needs an addition of 1% of Aliquat 336 (quartenary ammonium salts) to make it effective and have a relatively high viscosity to enable it’s use, this is called DIBK Aliquat 336 (1%) solution. It is made by having 10ml of Aliquat 336 placed in a 1lt volumetric flask and bringing the volume to exactly 1000ml with DIBK solvent. This is then the stock solution to make standards and direct use.
Oxidizing or non oxidizing solutions to be tested can be used with this method without any prior treatment but strong oxidisers eg. Straight Aqua Regia or any other nitrate containing oxidiser must have the nitrates expelled. Strong oxidizing solutions are more harsh on components in the nebuliser of the Atomic Absorbtion Spectrometer as well as being promotors of side reactions in the extractant components which can lead to false metal readings.

If the 10:1 concentration ratio gives an absorbance value which is higher than the value for a 5 ppm gold standard then a lower ratio is used.

A 1:1 ratio will just clean out interferents without enhancing the reading range.

If the solution you are testing has such a high precious metal level that a 1:1 ratio will still land you outside the 5 ppm limit, you can reverse the ratios such that the DIBK : liquor ratio is, say, 10:1 rather than the standard ratio of 1:10.

The reason you try to keep the metal readings in the DIBK below the 5ppm standard reading is that from 0 to 5 ppm the readings are straight line. This means that the absorbance reading for a 5 ppm standard will be 5 times as high as the absorbance reading for a 1ppm standard.

However, if you plot the readings above 5 ppm you will get what is called a calibration curve.

Most AAS machines have calculation circuitry in the electronics which will give you the actual value of the true absorbance reading for these higher values.

All of the above works brilliantly in a lab where clean synthetic solutions are being used, it tends to be more variable when real world gunged up solutions are being run.

The above represents the difference between an operator's idea of a bare bones AAS unit and the developer's idea. The developer's idea is that a simple unit only has relatively small amount of circuitry dedicated to the above type of calculations, the operator wants this type of circuitry kept to an absolute minimum to avoid processing errors.

The real difference is in the interpretation of the word "relatively", to the developer it seems to mean that he has restrained himself from installing as much software as he could possibly fit into the unit.

You can avoid most of the analytical problems just by keeping the DIBK/liquor ratios such that the absorbance values are maxed out around the 5 ppm standard values.

DIBK usually comes in 20 litre drums, Aliquat 336 comes in various 50 to 1000 ml containers.

The mixed DIBK/Aliquat solution is usually dispensed in 5 ml pump volumes from a dispenser mounted on a Winchester bottle.

Deano

 

[Quwannar]

Dear Deano,
You are a great man. Great and golden compiled knowledges for me. Thank you very much for your valuable sharings. I think that it has come the time to own an AAS equipment. Because it would be hard to tell or persuade the laboratories for a proper and a right analysis method required.
Thank you sir again.
Best Regards
Erdem

 

[Deano]

Many people have difficulty with the idea that any change in leach conditions will involve a trade-off.

The simplest gold leachant is cyanide using oxygen as the oxidising agent.

If you want to speed up the leach dissolution rate with physical inputs you can add heat or increase the amount of agitation applied.

Both of the above have limits above which the leach rate will not increase and both have input costs.

Basically you speed up the leach rate but it will cost you more to do so.

You are trading off time against money.

Similarly you can increase the leaching rate by adding different and stronger oxidisers rather than using the free oxygen from the air as dissolved oxygen in the leach solution.

These stronger oxidisers are usually peroxides of some form but exotic di and trivalent metals have been used.

The use of these oxidisers will increase processing costs directly but will also give problems down the line if you are zincing for gold recovery.

It will cost you more to remove the oxidiser from the leach solution to minimise redissolution of the zinced gold compared with using just dissolved oxygen from the air which is usually removed by vacuuming the liquor after filtration.

Here you are having an extra processing cost being traded off against time.

There are many cyanide based formulations available commercially. Most of these contains eclectic mixes of chelates and alternative gold solvents.

Apart from the extra cost relative to a straight cyanide leach you will have the problem of many other metals being leached by these added components.

In a zinc recovery from these leach types you will have to zinc out many of these extra dissolved metals in order to recover your gold.

In carbon processing you will have great difficulty in stripping the gold from the carbon, you are forced into ashing the carbon in order to recover the gold.

This is in addition to the loss of gold adsorption capacity on the carbon due to the loading of these extra metals.

In leaching some ores such as copper types you can minimise cyanide consumption by dropping the cyanide in leach tenor.

This makes the leach more specific for gold over copper but will slow the gold leach rate to the point where it can only be run as a heap or vat leach.

Here you are trading off cyanide cost against time and the inconvenience of having high copper levels on the carbon and thus in the carbon strip liquors even though you can minimise the copper levels in the strip electrowin liquor by performing a cold cyanide pre-strip on the carbon to remove most of the copper before attempting to strip the gold with a hot solution.

Usually when you are working with a high cyanide soluble copper ore you are trading off the lower cyanide consumption for longer leach times and more complex processing methods.

Pretty much all variations of all leaches involve tradeoffs of some form or another, the trick is in identifying the tradeoff and deciding whether you wish to go down this path.

Deano

 

[Quwannar]

Dear Deano;
Analysis result of 10 % HCL leach of silver slag is given below.
Head slag of sludge of table tailing below minus 150 mesh parrticle size.
Au 3,97 ppm
Ag 17.04 ppm
Filtrate of this table tailings
Au 31.8 mg/lt
Ag 370 mg/lt.
Pb 2715 mg/lt
Zn 817 mg/lt
Cu 94 mg/lt
As seen from results, the filtrate is more valuable part than solid sludge.
Can I recover the both Au and Ag together from liquor with saline/hypochlorite leaching method at pH 3 ?
Then can I load the gold and silver complex onto activated carbon?
İt is possible to sell the activated carbon loaded by mentioned metals.
Many thanks in advance
Best regards
Erdem

 

[Deano]

You need to have weights and volumes in order to tell what values are where.

What is the dry weight of the sludge and what volume of leach liquor as the filtrate.

You give the metals grade of the filtrate in mg/litre, check with your analyst to see if these values were milligrams/litre or micrograms /litre. It is kind of important as there is a thousand fold difference between the two.

Your recovery method depends entirely on which of the above grades is present, let me know the above values and I can advise you further.

Deano

 

[Quwannar]

Dear Deano;
I dont actually know the weights of sludge and the volumes of filtrate used by laboratory. They said that they could be capable of doing a standart DIBK/alliquat 336 extraction test after Aqua Regia digest. I sent a 75 ml of filtrate and 90 gram of sludge from which I run for a 200 gram slag sample insid 1.000 ml of 10 % HCl.
Units in report is ppm and mg/l. Milligram per liter
They did not state the weights and volumes.
Best regards
Erdem

 

[Deano]

What we need to know is the final volume of liquor from your leach and the final weight of solids after the leach.

If you started with 1 litre of solution, how much solution was left at the end of the leach.

Similarly how much solids was left after leaching.

I presume that the lab used aqua regia digest techniques rather than fire assay on both samples, if not please report which and why.

I am not worried how much sample the lab used.

Deano

 

[Quwannar]

Dear Deano;
I put the 200 gram of re-milled slag minus 100 mesh in 100 ml of 10 % HCl in one liter water. I completed the volume level with 10% HCL to maintain the loss due to evaporation. İn fact. I did not measure the volume loss and weight the sludge remained. And then, I picked up the sample mentioned amounts in my previous reply and I sent to laboratory.
So You can not do a math to help me about the yield and the chemical comsumption in the further leaching stage. I am so sorry for deadly mistake.
I am right now repeating the same expriment in specified wolume and the weights.
Expertise in lab said that you would do aqua regia digestion for especially in sludge. But you dont know what to do with the filtrate. I remembered your instructions to them about the filtrate. Lab guys prepared 1;1 volume filtrate and aqua regia. And did the dibk and alliquat 336 extraction and AAS reading.
Thank you very much for your valuable time and your efforts.
Best regards
Erdem

 

[Deano]

You are trying to find out several things from the HCl leach.

1. What % of the slag is dissolved in the leach.

2. What % of the precious metals reports to the leach residue and the leach liquor.

3. What is the grade of the precious metals in the residue and liquor.

The above findings will show you which products of the leach process are viable to attempt to recover the metals from and the likely return from each product.

These results may affect the leach type to be employed in commercial processing and so are very important.

Usually once you have the indicative numbers from the leach test, if they appear viable, you would do multiple repeat tests in order to get an average grade of possible recovery for the whole deposit.

Always keep in mind that the numbers will vary greatly from sample to sample depending on the grade and processing of the original smelting.

Your first numbers, even though not related to known volumes and weights, appear to show the slag contains the high precious metal levels which are to be expected in the slag.

Usually three things will happen in relation to the processing of this type of material.

1. People will take the head slag and apply a standard assay analytical technique without first doing the HCl leach.

The results will be low and disappointing and you will be discouraged.

2. People will look at the results from the HCl leach products and decide that the numbers are too high and that you have a flaw in your test regime.

Unfortunately these people are often big wheels in processing and their pronouncements will carry a lot of weight and once again you will get discouraged.

3. Someone with a vested interested will propose an alternative method of processing and when it fails will claim that the grades were not there in the first place.

The best way to negate all of the above is what you are doing, namely simple repeat testing with independent analyses.

It is important that all of your testing is recorded in detail so that you have all of your numbers available for review if needed.

Finally, if you carry out the HCl leach on just the table cons by themselves, they will clean up to the level where they can be dried and smelted to give you a silver bar with low but very important gold levels.

It is important that the smelt is done without flux in a clay crucible otherwise you are going to restart the entire metals in slag process.

Usually you put the clay crucible inside a graphite crucible as a safety precaution so that when you are reusing the clay crucible you do not get too worried about any breakage of the eroded clay crucible.

Keep in mind that the metals from the table con will usually be the majority of the metals recovered so it is worth doing the processing correctly.

Deano

 

[Quwannar]

Dear Deano
Our Master;
I do understand the sipirit of your statements well about the proof of precious metals entities in slag by instrumental analysis. I could not still persuade lab guys working in highly prominent labs. Everyone does not want to give up their the standart comforts. Thats their culpa maxima.
As you brillantly propose, the coming back from the outcome has always been useful tool in the science and art of gold and its engineering.Because man do want to see the gold.
The method you developed is a some kind of a revolution in slag processing. And your posts in GFR must also be lectured in faculties of engineering.
Meanwhile, I smelted 40 gram of dirty concentrate containing some fayelite with some dry borax. Smelting conditions was not suitable for full needs. I saw real gold disseminated at the slag surface.
Knowledge and wisdom be upon you sir.
Best regards
Erdem.

 

[Deano]

In analysing digest solutions for precious metals by AAS using DIBK / Aliquat336 as an extractant there is often a gel formed in the organic phase which prevents aspiration.

Most users of this technique have their own preferred method of breaking this gel phase so that a liquid sample is available for aspiration.

These methods usually employ an acid of some form, phosphoric and hydrochloric are among the most common.

The gel is usually formed by the presence of micro particles of silica from the digest and the best way to break the gel is by the addition of Hydrofluoric acid to the DIBK after phase separation.

I do not know of anyone who is really comfortable with using hydrofluoric acid as such, sooner or later there will be spillage or droplets onto the skin or gloves.

A much safer alternative is to add gently some ammonium bi fluoride powder to the gel, the dissolution of the silica is rapid and it is not necessary for all of the gel to be broken, only enough for aspiration to occur.

The AAS reading for gold is totally unaffected both up and down by the addition of up to 1 gram of the powder to the gel.

Deano

 

[Deano]

When I started in the gold game last century the recycling industry had not really started in regard to e waste.

It was a case of learn as you go with large quantities of scrap.

The first thing I learned was to use only cyanide as the gold solvent, any acid based methods were either too expensive or metallurgically tricky.

If you were using alkaline cyanide solution it made sense to remove any coating metals with alkaline solution so that you did not have to go through a neutralisation step which was costly when running large quantities of material.

I found that tin, zinc and lead could be cheaply, easily and completely solubilised using sodium hydroxide solution.

You had to play a game of time and temperature versus removal efficiency of the metals.

At 1% caustic solution you needed to have the temperature up around 80C for good metal removal.

At 20% caustic you could get good removal at room temperature.

Temperatures and caustic strength were pretty much straight line between the two extremes.

I generally went for the 20% option as there were no heating costs.

The thickness of some of the metal coatings meant that some agitation was needed, I just used a peristaltic pump to move the liquor through the material, this meant that any reactions with pump components were eliminated, you just needed an alkali compatible hose.

As there was no sorting of components in those days the time of the alkali leach was controlled by the leach time of the thickest coating, could be up to 24 hours.

All dissolved metals were removed from the leach liquor by running the liquor through a column of activated carbon and the caustic level adjusted before recycling the liquor.

If you are looking to use acid based methods for the gold dissolution step it is not hard to do an alkaline leach, rinse and then come in with the acid leach.

Deano

 

[anachronism]

Dean

Did you put the liquor at 20% Caustic through the carbon or did I miss your point? If I recall correctly from what you've told me an alkaline solution loads to a lower degree on carbon than an acidic one- could you please clarify?

Jon

 

[Deano]

Good to see that your memory is as good as always.

Yes an alkaline solution will generally load to a lower degree than an acid equivalent solution will, but both will load to a level which meets the requirement of removing the metals from solution.

Consider the two solutions to be the equivalents of a Model T Ford and a Ferrari, both will get you to your destination without all of that tiresome walking but the Ferrari will get you there faster.

With base metals it is best if any heated solutions are cooled before treating with carbon.

Deano

 

[METLMASHER]

:shock: Great reading, and my heartfelt thanks to Deano and the GRF

 

[Deano]

Heap leaching

One of the many myths about heap leaching is that any recovery below the theoretical amount is due to the mineralisation being locked up in matrix.

Occasional mention is made of possible channelling of the liquor flow but this is usually relegated to a minor secondary position on the “why did my recovery not reach near optimal” scenario.

If the leached heap is pulled apart so that the internals are fully exposed, a totally different pattern emerges.

Large areas of the heap, average working volume 10% of the heap, are totally unleached.

In a copper heap leach the difference between the whitish leached areas and the green unleached areas is stark while the sheer volume of unleached material is stunning.

The unleached material forms large zones, many house sized, in a variety of sizes and shapes.

It is obvious that the leach liquor has not performed any leaching in these zones.

As well as these unleached zones inside the heap there are other unleached zones around the perimeter of the heap.

One of the other myths regarding heap leaching is that the leach liquor will travel substantial distances horizontally in the heap.

This is incorrect and the liquor flow appears to be controlled more by gravity than any other factors in a heap leach.

This means that there is an unleached zone of ore which starts approximately two metres in from the outer edge of the top of the heap and will include all ore from the line starting two metres in from the top edge to the outer edge of the heap.

The two obvious questions in relation to the unleached zones are;

Why are they there

And is there a way to leach them without having to remodel the heap entirely.

Unleached zones are purely an effect of using sprinklers to distribute the leach liquor onto the heap.

Apart from the sprinkler side effect of depositing gypsum on the top of the heap due to excessive evaporation and thus forming an impermeable cap which prevents liquor penetration occurring evenly across the top of the heap, the liquor flow from a sprinklered heap tends to run through the heap in defined channels.

This is the cause of the unleached zones in the heap.

The size and number of these dead zones is also dependent on the ore type, milling and placement protocols.

There is also no sideways distribution of the leach liquor in the heap so the leaching start line is vertical from the inside edge of the top bund wall.

The unleached zones within a heap may be leached by changing the liquor introduction method to the heap from sprinklers to flooded paddocks.

This alters the liquor flow patterns inside the heap so that the flow now goes through the previously dead zones.

In the paddock method the bund walls are made around 1m high by 2m wide, no compaction is employed.

This allows the liquor flow to move sideways through these walls and thus gain the two metres horizontal distance before distributing vertically.

Before the bund walls are formed the top 0.5 metre is removed from the top of the heap. This removes the gypsum layer from the sprinklers and allows better distribution of the liquor through the heap.

Depending on the flow rates through the heap and how the treating circuit is designed and sized, a paddock size is often around 100 x 50 metres.

The surface is levelled before the bunds are constructed.

A vital part of the surface conditioning is the compaction of the surface so that a 300 to 400 mm depth of water in the paddock will take around 8 hours to drain dry.

After the liquor has drained from the paddock the paddock is lain idle for an hour or so to allow air to be drawn down by the liquor to access the heap’s interior.

The paddock is then refilled and the sequence repeated.

Fill time is usually around 1 to 2 hours, scouring of the surface is prevented by introducing the pumped liquor into the paddock via a tangential pipe entrance into an old truck tyre with the liquor spilling out via the centre of the tyre.

A further major advantage of the paddock method is that the alternate cycles of liquor flow through the heap followed by the air access to the ore will greatly speed up the degradation of sulphides present in the ore particles, this applies for both acid and alkaline leaches.

This sulphide degradation will allow the leach solution access to more metal values in the ore particles.

Deano

 

[snoman701]

Hey deano....

When running a low concentration cyanide leach how do you monitor cyanide levels in the leach? Let’s say you have a target of maintaining 10 ppm free cyanide.

Are you just doing a particular titration, or are the meters good to that level when calibrated against good standards?

Does the use of a ferrocyanide leach complicate monitoring since free cyanide could be low when total cyanide (free cn+ferro-cn+aucn) is otherwise appropriate?






Sent from my iPhone using Tapatalk

 

[Deano]

The handheld meters have pretty well relegated titration to being used just as a regular check on the calibration of the meters.

The meters are good to extremely low levels of cyanide, the usual problems are lack of care with the probe, especially leaving the probe in solutions of free cyanide, this will affect the zero point and also allowing the probe to dry out.

All that you are interested in is the free cyanide level and that is what the meters will register.

If you are using the ferrocyanide leach in sunlight then pretty well all of the conversion to free cyanide will have occurred.

You are only interested in the free cyanide level when trying low cyanide gold preferential leaching, especially with copper present.

Deano

 

[snoman701]

The handheld meters have pretty well relegated titration to being used just as a regular check on the calibration of the meters.

The meters are good to extremely low levels of cyanide, the usual problems are lack of care with the probe, especially leaving the probe in solutions of free cyanide, this will affect the zero point and also allowing the probe to dry out.

Is the titration against silver nitrate with potassium iodide the best method to check calibration? Or is there a better version?

What brand meters (probe) hold up well?


Sent from my iPhone using Tapatalk

 

[Deano]

Titration testing is done to check for gross errors in calibration of the meters or if it is suspected that the cyanide standard used has degraded.

You usually use the same cyanide source in calibration as used in the plant, If that source has degraded then a new check source is used.

Pretty well all sources of cyanide are reliable in regard to freshness of supply, this was the most common problem with cyanide stored in mine sites.

Most probes come out of China and are surprisingly reliable if looked after.

Buy by price, most of the meters also come from China and are similarly reliable.

Deano