Monday, 6 March 2017

Flotation Chemistry vs Flotation Physics

I have had an interesting email from a very eminent flotation researcher, which I am sure will generate some debate. He/she would prefer to remain anonymous as he/she feels that industry has a reputation for judging the whole department by any individual. I publish the email below with only minor edits:
Question – currently, why is so much attention (≡ to money) being spent on the comparatively narrow study of the physical aspects of flotation, at the expense of flotation chemistry studies? A good example is the current preoccupation with coarse particle recovery.   
As a starting point we should ask ourselves - ‘What is flotation all about, really’? The answer is straightforward - the ‘Holy Grail’ of flotation is selectivity not recovery. After all if we take an ore body and do nothing we have 100% recovery, but the product, as such, is useless. Recovery without selectivity is all in vain. However there seems to be a meagre acceptance of the importance of grade.

Little is said about selectivity but there is a constant cry for increased recovery. This might be what industry wants but it is certainly not what industry needs. However, physics based studies are merely concerned with increased recovery not increased selectivity; selectivity in flotation is only achieved by attention to the pulp chemistry. Yet, currently, there appears to be little (research) money being directed towards chemistry based flotation studies. Strong, and successful, chemistry based research groups like those at CSIRO and The Wark have all but disappeared because of a lack of (financial) support. 
As natural resources become more and more scarce, and more and more complex, their value will be determined more and more by the degree of selective separation which can be obtained in processing. This can already be seen for arsenic contamination. And selectivity against pyrite will increase in significance, exponentially. So, we might well ask if the current strong financial support for the physical aspects of flotation is misguided. 
For example. An actual case study - the treatment of a sulphide ore. For the rougher stage, recovery beyond 85% was difficult. The rougher concentrate, at 85% recovery, could be worked up to a smelter acceptable product of 25% grade; the tailing assay was 0.09%, with most of the loss in the coarse (-500+40 micron) size fraction. With some effort (in a laboratory) it was possible to do a size split (at 40 micron), retreat the coarse fraction, and recover another 10% of the values in a scavenger product (ie lift overall recovery to 95%). Assay of this scavenger product was about 1%. This gives a ratio of concentration of about 10, similar to what was (and is commonly) achieved in the rougher stage. However, no amount of retreatment could raise the assay of this product above 6%, and this grade is not acceptable at the smelter. Accordingly, in practice, these particles, after initial recovery, would have ended up back in the tail. Therefore, in this case, any money spent on installation and operation of special cells for recovery of these coarse particles would be misguided.  
Second (obvious) question. How many of the researchers seeking special ways of increasing the recovery of coarse particles (for example) have thought of what they are going to do with these particles when they are recovered; perhaps the final recovered value is less than the cost of recovery? 

Third more important question. How many of the people putting up the money for this research even understand the problem? And would some of the currently available research money be better spent on further understanding the problems of, and developing techniques for, upgrading low grade concentrates?
Flotation chemistry studies are certainly more difficult than physical flotation studies; what is not generally recognized is that they are also more important.
And all this prompts the last question – does anybody really care?

Strong opinions and I invite further strong opinions! Personally I found "...but there is a constant cry for increased recovery. This might be what industry wants but it is certainly not what industry needs" contentious. Increased recovery may or may not be what industry needs. But what about society? As we move towards a circular economy isn't maximum recovery of natural resources, at optimal grades, what we should be aiming for? And is there more to increased recovery at optimum grade than just flotation? Comminution-liberation must play a big part in this (but that's another story). Hopefully we will hear much more about this at Flotation '17 and Comminution '18.


  1. When times are hard, funds will go to physical flotation as it offers more immediate potential for payback. Flotation chemistry is seen as more arcane and can be ignored through trial and error giving the optimum reagent combination. Why or how it works is a luxury which is pursued when times are good and money can be spent on the full optimisation.
    As usual, industry will only support things which will lead to short term goals bejng met.
    The death of specialist groups will not concern industry one jot, they will simply carry on as usual, blundering around in the dark until a solution is found for the speciffic problem at hand.

  2. Strong words indeed, but the first author misses a major point - most collector reagent schemes currently used in sulphide mineral beneficiation have been known for long times, and they work fairly well. Each has its limitation, for example xanthate likes chalcopyrite and all disulphides, but this can be alleviated by changing to a dithiophosphate regime. Then, if some miners do not understand this, its their problem. And, of course, you need to grind to a fineness so the bubble may lift the particles out of the pulp.

    What I find lacking in sulphide flotation reagent research are:
    * Poor understanding of gangue depressant's action, and links to their structure;
    * Poor understanding of the interaction between collector and frother, especially froth stability in large tank cells;
    * Add your favourite ---

    Some of the points above have been addressed by manufacturers of reagents, for example Cytec has developed polymeric reagents for gangue depression, but I haven't seen much academic research in this area.

    However, I believe that the major chance to develope new reagent regimes is in oxide and silicate mineral flotation. I see a need for:
    * Collectors that may float Ca-mineral selectively (Apatite-Calcite);
    * Less toxic alternatives to replace amines in silicate flotation.

    So, there are research challenges if you dare to check for them.

    Bertil Pålsson
    Minerals and Metallurgical Engineering
    Luleå University of Technology

    1. Flotation is not a panacea for all ills.
      Look wider. Only in this case you will have success. There are many methods of separation (concentration) of minerals:
      Gravity separation (There are more than 12 types ).
      Magnetic separation (There are more than 14 types).
      Electrical separation (There are 6 types).
      Flotation separation (There are more than 17 types).
      Special methods of separation (There are more than 15 types).
      Chemical separation (leaching) (There are 7 types).
      Also, there are many combinations of methods of mineral separation:

      •Flotogravity separation
      •Magnetohydrodynamic separation
      •Pyroelectric separation

      •Tribo-aero-electrostatic separation
      •Pneumoelectric separation
      •Fluidising-electrostatic separation
      •Crown-magnetic separation
      •Opto-electrical separation
      •Segregation-diffusion concentration

    2. Yes Natalia, all those methods have their place, but it is impossible to over-emphasise the importance of flotation. Around 20 million tonnes of copper, arguably the most important base metal, are mined annually and virtually all this tonnage, if not all, is concentrated by flotation. As I have said before, it is the most important technological development since the discovery of smelting

  3. I dont have a problem with researching coarse particle flotation as some ores will benefit from it and some won't. Let's understand its applicability. Coarse particle flotation is not about concentrate grade, it is about tailings grade. If you can't recognise that then you have missed the point. The holy grail of any recovery-targeting separation is to generate a tail that is throw-away grade. If that hasn't been achieved then there is an unacceptable revenue loss and the technology is not applicable. When successful, coarse particle flotation will be a wonderful energy minimisation tool that will convert some deposits into orebodies.

    There are many other neglected research areas, such as froth washing, which are purely physical and offer amazing benefits. I have seen massive froth washing benefits ( like generating +30% Cu concentrates with washing when detailed flotaton chemical studies couldn't get past 23% Cu) so I can't understand anyone wanting to concentrate on one aspect of flotation at the expense of the other. Let's keep up pressure and justification for looking in detail at both chemical and physical aspects of one of the most remarkable mineral separation tools we have.

    1. What type of ore did you write about? How many types of ores do you know? Do you understand that this advice is suitable only in a particular case?

    2. Natalia, I work with all ores, flotation recoverable or not and have been doing so for 35 years. Commenced in the 80s at Broken Hill on Lead/silver/zinc, developed flotation for gold/pyrite ores, mineral sands, magnesite, graphite and chalcopyrite from galena. I also ran flotation pilot plants in all but one of these (and all that just the first 4 years after graduation). I developed mineral sands flotation further at project level then went on to diamond processing (Argyle). Even investigated diamond flotation and saw 4 mm stones genuinely reporting to concentrate by true flotation in Siberia. Since then I have consulted globally in comminution and physical separation (including a lot of flotation) as a JKTech consultant and, for the last 14 years, in the engineering and design field. There are few flotation ores in the hard rock mining business I have not worked with directly. There are certainly a number of ores in the industrial minerals world that I have not worked with.

      All rougher flotation for all ores is about recovery and all cleaner flotation for all ore is about grade. All too often I see comparisons of rougher performance based on the grade achieved - this is not relevant as long as there is a good mass rejection. To reduce comminution cost this must be done as coarse as possible and this is directly driven by texture and liberation. Up until a few years ago coarse has meant 200 - 300 um P80 (for example in porphyry copper). The current coarse flotation work is looking at ~1 mm range. If 50% of the mass can be rejected at 1 mm with only a couple of percent loss of metal then excellent (note that it is still a work in progress and I am not directly involved so have no promotion agenda here). If the process is practically implementable at large scale then it will mean a few kWh/t will be saved by not grinding half the plant feed, but its success is entirely driven by the cost benefit analysis on each particular ore.

      The ore I described in my first post was a fine grained IOCG copper ore and about 70% of the value liberates at 40 um but the remaining 30% is very fine grained and requires a 12 um regrind. The ore also contains pyrite, which in turn, carries liberatable gold. Copper/pyrite composites float in the rougher stage (75 um P80) and pyrite is deliberately collected in roughing to maximise copper and gold recovery. An initial regrind to 25/30 um allows 70% of the copper to be recovered in a high grade concentrate while all pyrite and Cu composites are rejected to coarse cleaner tails. A further 12 um regrind on the tails releases the remaining copper (90 to 93% recovery overall) and liberates half the gold that was locked in the pyrite (the remainder is sub micron gold inclusions confirmed by LAICP-MS investigations. It is the discrimination between copper sulphides (mostly Cpy) and pyrite at the 12 um and 25/30 um cleaner stages which benefit strongly from froth washing. Pyrite is not bound in the froth and falls out with washing while there is no measurable loss of copper in concentrate. This is a classic example of where the physical and the chemical are both varied dramatically through the flowsheet to maximise both grade and recovery while minimising both OPEX and CAPEX. All of this flowsheet was proven in the laboratory using conventional cells (Denver/Agitair) and then conventional cells were compared with a washed column operating in parallel at pilot scale to arrive at the comparison I quote in my first post.

      This is a specific case, but I apply equal weight to the physical and chemical in all flowsheet development work I do so that the best possible outcome is achieved in all instances. Process mineralogy and liberation analysis is the starting point and, once I (and my colleagues) have listened to what the ore has to say to us, the plan to develop the flowsheet follows.

  4. I think everyone including the original writer has made very good points. However, the business of flotation and any other unit operation considering a separation, needs to consider the economic viability of a process. A large proportion of cost is dedicated to trying to liberate minerals to the extent that flotation will be an appropriate separation process, if we can float at coarser sizes (even if it is less liberated), it benefits the overall economics of the flotation process tremendously. That being said, the separation still ultimately relies on applying the appropriate chemistry, and there is plenty of research to be done especially around depressants/dispersants (as mentioned in one of the other comments), as well as furthering our understanding of reagents and how they perform under different water chemistry conditions.

  5. This question is difficult to answer and exactly why flotation is referred to as a physicochemical separation. The chemistry is a balancing act of solution and surface chemistry, predominantly for frothers and collectors. The physics is a balancing act of particle/bubble sizes and hydrodynamics. With over 50 parameters to be concerned with, flotation is indeed an art/science and we are only now getting it to be truly engineering.
    Courtney Young, Montana Tech, USA

  6. I would like to look at the question in a simple manner.. For me, maximum recovery at a given grade is a must. This depends on liberation and the size distribution of feed to flotation.If one is lucky with his ore characteristics, one may get maximum recovery at a grade higher than wanted by metallurgist; then one can blend lower grade concentrate obtained at higher recovery with the earlier higher grade concentrate.
    But water also has a role to play. One should look at fines going into concentrate/tailings and their mineralogy and liberation Is desliming before and after flotation necessary?Is regrinding of one of the streams necessary?
    I feel there is enough said on physics and chemistry of flotation--let us look deeply into the behaviour and characteristics of particles moving around in rougher/scavenger and cleaner circuits.

  7. It is time to move beyond the reductionist breakdown of flotation system that has served so well in the past and to progress with a more balanced approach such as suggested by some researchers, i.e., jointly considering particle characteristics, chemical mechanisms, and physical mechanisms. As a comment above mentions this is a physicochemical process working on a highly variable feedstock of natural origin.
    The differential development of hydrophobicity is definitely a critical sub-process in flotation. However, it is difficult to argue in the realm of sub-processes that it is the single one deserving of focus to the exclusion of others. Consider the journey of particles from orebody into the froth concentrate:
    • Breakage dependent on orebody geology and mineralogy and comminution process leading to liberation profiles and surface characteristics which set the stage for interacting with collectors, depressants, activators, frothers, etc. and development of a hydrophobic-hydrophilic character
    • Surface reactions leading to hydrophobic surfaces or not
    • In the pulp - bubble-particle interaction, collision, attachment, detachment, bubble-particle aggregate migration to the pulp-froth interface, particle elutriation to the pulp-froth interface
    • Bubble-particle transition or not across the pulp-froth interface, bubble-particle migration from the interface to the launder, particle elutriation across the interface and transport into the concentrate
    Should surface chemistry be the essential focus for process understanding and improvement with this knowledge? As further considerations:
    The importance of orebody characteristics, chemical mechanisms, and physical mechanisms.
    • Variability of ore as seen in geometallurgical analysis of orebody and lack of transparency in relationships between geological and comminution, flotation, and solid-liquid separation characteristics.
    • Is the low froth recovery observed in flotation systems, particularly for coarse particles, of physical or chemical origin?
    • Is the size-recovery behavior of chemical or physical origin?
    Bob Seitz

  8. You need to look at both of them together, flotation physics and fotation chemistry. Flotation is a physical separation process but the attachment to the bubble is enhanced or affected by the chemical reactions in the slurry. Ultrafines are lost to tailings mainly due to flotation cells limitation. I would like to point out that selectivity and recovery are both important in flotation because you may reach recoveries in the range of 97% with very high selectivity so please do not dissociate and pick either selectivity or recovery ; similarly do not go looking just for hydrodynamics of flotation but also look at the chemistry at the same time. Please don't forget to include the noise introduced by operators and by control systems. Groups working to look at the flotation chemistry should not disappear but should be enhanced by asking them to include physical and chemical variables . One more thing, don't forget to include mineralogical studies in your research.
    Juan J. Anes, EM2PO, Canada

  9. We have to look at it in a simple way==if the liberation is at coarse particle size, one should not even think of flotation which is expensive and environmentally nor accepted; go for gravity separation.
    There is only adsorption of the chemical on the particle and there is no chemical reaction. Bubble gets attached and makes up the total volume(i.e.particle and bubble) which in turn makes the total density less than water and floats.
    We have to realise that each particle of feed to flotation has its own size and specific gravity depending on the minerals content in that particle. It then follows that we have to give a bubble(s) of enough volume to get attached if we want that particle to float.

  10. Both. When developing the flow sheet for a new deposit / project "known" reagents / chemical conditions are used in the test work (to enable ready comparison), yet conversely when the plant is built and operating the focus switches to developing 'bespoke' reagent blends and optimising reagent addition points / dosages (I include aeration as a reagent).
    Chris Smith, Rio Tinto, Australia

  11. I suggest you read my book "Fundamentals of the theory of flotation"
    You will understand very much.

  12. An interesting question Barry . . .

    There are two fundamental principals that guide the success or failure of all industrial flotation. They are: are the minerals adequately liberation; and do the minerals of interest have the right surface chemistry (i.e. hydrophobicity). Without liberation the grade of the concentrate produced will be below par. However, if liberation is adequate and the surface chemistry is not right then it is unlikely that the desired grades and recoveries can be achieved.

    So, in the first instance, in a real system along with liberation the chemistry of the system is the dominant parameter that controls the success of failure of a separation by flotation.

    No amount of tinkering with the physics of the system will significantly incease the probability of a particle of interest actually contact a bubble and remain attached if it does not have sufficient hydrophobicity. Understanding the chemistry helps with they physics.
    Chris Greet, Magotteaux, Australia

  13. Well said,Chris.
    By using "highphy"terminology, we made flotation look as "rocket science"; break the process into mineral characteristics and liberation followed by how to make wanted particles to float by attaching them to air bubbles after making their surface hydrophobic.

  14. Flotation chemistry Vs Flotation physics? Is it a good question? Both aspects are important, is it not?


  15. (1) Thinking of flotation chemistry, have we gone beyond what is said by Gaudin and Glembotskii (their text books)? Research in flotation chemistry used to be about sulphide minerals with focus on copper activation and xanthate flotation of sphalerite. Though di thio phosphates are known to be more selective in sulphide mineral flotation, tests/trials in Indian plants did not produce unequivocal results.
    (2) The problems of complex lead zinc ores of India which I know decades back are still there for example effective depression of graphite/mica gangue. Nigrosine black is an effective depressant and a better replacement could not be found. Silica reporting into zinc concentrate. Recovery of galena particles finer than 20 microns. Recovery of silver etc.
    (3) Which fatty acids are more effective in oxide, silicate and salt type minerals? Saturated fatty acids or unsaturated? What does the theory of flotation say?
    (4) In my opinion the research of Prof. JM Pratt and his group during eighties is cut apart. They reported that pure sphalerite floats (a) in urea solutions, (b) upon activation by Cu ions in the absence of xanthate (or any other collector). They also reported that ‘change in flotability is associated with a change in the structure of water at the interface’.
    (5) During IMPC-2012, JD Miller group presented experimental proof that adjacent to hydrophobic mineral surface there is ‘water exclusion zone’ causing change in the water structure. The research mentioned in (4) & (5) may be of interest to those interested in the theory of flotation.

    DMR Sekhar

  16. in my view flotation physics and chemistry go hand in hand.
    The liberation of the particles which is obtained by size reduction and classification is unit operation.Here the physical properties of the minerals play a vital role like mineral hardness and mineral density. The mineral particles which are ready to the separation must have suitable size for responding to the separation technique. For example in flotation, the finer hydrophobic mineral particles have less probability to have fruitful collision with the bubble in the dynamic environment in the flotation cell. other hand relatively coarser particles have more probability to have a fruitful collisions with the bubbles. Now the retention of mineral particle on the bubble will be dependent of the hydrophobicity(unit process?) of the particle. Coarser particles >500 microns might not have sufficient hydrophobicity over coming their drag force due to their density. So, there might be a low chance for the upward movement of the coarser particles in the froth phase.

    1. Dear Sir,

      The probability of fine particle (say less than 20 microns) colliding with the air bubble is less, agreed. What can be done? (1) increase the RPM of scavenger cells and (2) finer particles demand more collector. Use collectors with longer hydro carbon chain. Fine particle flotation is to be looked at through both from chemical and physical techniques.

      DMR Sekhar

  17. Good remark,Dr.Sekhar. Let us have cyclosizer data on these fine particles. Depending on chemical analyses of these fractions, we may even think of using cyclones/stub cyclones to separate extreme fines
    Very interesting discussion.

  18. Comminution-liberation must play a big part in this (but that's another story).
    I will be very interested in this topic. Most of the losses in industrial plants are due to problems in the grinding area. The optimal particle size or "liberation" for type of mined ore is the major disturbance these days. In addition, the ore containts large amount of impurities and waste that is treated. This is where I spend a lot of time. We using Machine Learning Algorithms to find the best operating strategies. I am amazed at the results. The computer "learn" from the grinding and flotation data. It is a very interesting topic. This is not new. We have published several papers on this topic. It just became easier to work with. (Copper 1991, Integrated Grinding/Flotation Controls and Management). The algorithms are mostly free. The strategy is to get the right data.
    Copper 1999

    We need to minimize the processing of unuseful materials by conventional grinding and flotation. Just a comment for further conversation.

  19. Osvi Bascur,

    I thought that the discussion is about to close. As pointed out by you and Rao Sir, prevention of fine grinding of sulphide minerals is the key.

    DMR Sekhar

  20. (1) Bertil Palsson:
    Interference of floatable gangue (graphite, mica, talc etc) is indeed a problem. Dr. K. Ravindranath successfully used Nigrosine Black as graphite mica depressant decades back and it is still being used in the lead zinc plants of India.
    ‘I believe that the major chance to develop new reagent regimes is in oxide and silicate mineral flotation.’ True. The consumption of soap collectors can be reduced substantially if soap is accompanied by a small dose of ‘Alfa Olefin Sulphonate’. Soaps that contain straight chain fatty acids (up to 30%) are more efficient. This is a plant practice in India. The problem with saturated fatty acids is the solubility.
    (2) Dean David:
    There are many other neglected research areas, such as froth washing, which are purely physical and offer amazing benefits. Agreed, Sir.
    (3) Nataliya Petrovskaya:
    ‘I suggest you read my book “Fundamentals of the theory of flotation”. Thanks, Dr. Nataliya where can I buy that book?
    (4) Osvi Bascur:
    ‘I will be very interested in this topic. Most of the losses in industrial plants are due to problems in the grinding area.’ But how to prevent liberated chalcopyrite, galena & sphalerite particles going back to the ball mill via the under flow of hydro-cyclones? What about using fine screens? A few decades back Prof. TC Rao was talking about ‘double classification’.
    Thanks to all those who posted comments here. Thanks to Barry for initiating the discussion.

    DMR Sekhar

  21. HI DMK Sekhar:
    Thanks for your comment. Your suggestion said that we need to invest more in the classification methods used in the grinding circuits. We have looked at some strategies. We need to look at this depending on the commodities. In general, just having the right liberation and size distribution for maximizing flotation recovery is the major goal.
    And Yes, I remeber about Prof Rao comments. I have used his modeland improved for Optimill and Dynamill.

    1. Dr. Bascur,

      Incorporating 'unit cell' in the grinding circuit may also of of use. 'Practical understanding and application of the effect of dual outlet operation on industrial flash flotation cell hydrodynamics'

      DMR Sekhar

  22. Barry,
    I like the discussions on this topic.We got wonderful comments and they all bring out the real aspects of flotation practice--so many other aspects are indeed interlinked with flotation.
    . Perhaps this might be the topic on which you could generate so many remarks.
    Keep such lively and relevant topics to float around--researchers would get enthused and will look at each unit operation in mineral engn holistically.
    Thank you.


If you have difficulty posting a comment, please email the comment to and I will submit on your behalf