Mineral Processing

Note: The context of the process mineralogy data requested and supplied should be very clear, and needs to represent the deposit/material to be processed as closely as possible. Mineral processing is performed at many tonnes per hour, and for years, so any information obtained on small and non-representative samples is often useless. In the development and operation of mineral processing plants, large economic decisions are based upon mineralogy, and if the samples provided for study are not representative of potential run-of-mine ore, flow sheet and process development will be less than ideal and will lead to long term loss of recovery and expensive mill retrofitting.


Coarse-grain placer monazite (notice scale @ 500um)


Residual primary bastnäsite grains with web-like secondary REE minerals being produced from them due to weathering.

Below are a few questions that need to be addressed in the process mineralogy studies, if not already addressed in the geological mineralogy work:

  • What are the REE-bearing minerals, and how much of each exists? Examples are bastnäsite, monazite, xenotime, synchysite, allanite, etc.
  • Are there any possible economic by-product minerals? Examples are barite, fluorite, pyrochlore, columbite, uranium and lithium minerals, etc.
  • Are the REE minerals a by-product? REE bearing minerals can often be a by-product of other process operations. In these cases, REE recovery will be directly affected by the operating company’s interest and relative economics.
  • Are there both minor and major hosts of the REE? Examples of minor REE-bearing minerals are apatites and zircons—these minerals can often contain a few tenths of a percent REE.
  • What percentage of each REE is carried by each REE host mineral?
  • Are there any silicate-based REE minerals? These minerals are known for difficulty in the hydrometallurgy stage.
  • Are the REE minerals primary or weathered/altered (also known as “supergene”) Historically, weathered/altered REE minerals are difficult to process due to extreme fine grain size, needle or dendritic-like structure.
  • What is the shape/structure of the REE minerals?
  • If reasonably round, what is the average size and size distribution of the REE minerals?
  • After grinding, what is the % free REE mineral surface area as a function of particle size? i.e. how much REE mineral liberation exists as a function of particle size?
  • What is the chemistry/assay of the REE minerals? i.e. based upon microprobe/equivalent analysis what is the % of each important element in the REE mineral? Having this information will allow an estimate of the grade of potential mineral concentrate.
  • What are the gangue minerals, their amounts and characteristics?
  • Are there any impurity elements of concern, and in which minerals do they reside? i.e. thorium, uranium, arsenic, lead, etc.

Some definitions regarding Mineral Processing:

MINERAL LIBERATION – The relative amount of surface area, after grinding, of a particular mineral grain. For example, a 30% liberated bastnäsite grain means that 30% of the surface area of that grain is bastnäsite and 70% is something else. The something else is one or more other minerals. Mineral liberation is essential for the mineral processing engineer, in that high grade mineral concentrates cannot be made without sufficient amounts of REE mineral liberation. This mineral liberation is produced during size reduction procedures (such as grinding) where finer grinding usually means increased liberation.

MINERAL PROCESSING SCOPE – The development of a REE processing flow sheet is composed of three major stages, with complete success and understanding required at each stage before the next stage is attempted:

  1. LAB SCALE – The first is lab scale batch tests, also referred to as scoping studies. The sample amounts per test are in the gram to a few kg range. These are a snap shot of a particular lab procedure on a particular sample, which may or may not be representative of the bulk geological material of the mineralized region in question. Often the best (and easiest to obtain) samples are tested first.
  2. PILOT PLANT – The second stage of lab testing is termed pilot plant scale. It is continuous, and intended to be scalable to full production. It utilizes somewhere in the range of 10-100 kg per hour, and runs for days/weeks and includes a variety of potential ore types. Engineering data for feasibility studies and scale up is obtained during this process.
  3. PRODUCTION – Full or production scale procedures are the industrial equivalent of the pilot plant, continuously operated with 50 to 1000 tonnes per hour as feed.

PROCESS RECOVERY – The percentage of the REEs that show up in the product or midstream process relative to that of the mined ore or previous process step. Results are usually given in % values. Typical values are in the 50-90 % range.