Direct fire assaying after pulverizing --please note sample preparation is the core factor a rotary splitter must be used.

Usually the assay results from carbon are highly variable. This is due to the heterogeneous nature of the material (carbon and fine sand, silica) and the different densities of the components in the sample leading to segregation of the sample.Therefore pulverizing will mix the sample homogeneously.

Carbon is almost universally ashed prior to fire-assay. This is because the carbon is a strong reducing agent and will reduce all the litharge in your flux resulting in, well, a mess. Ashing reduces the carbon sample to a manageable size. If you want good results, ash a large number of samples and compare the variability of the results. 

Check out Bugbee - A textbook of Fire Assay. There are excellent practical guidelines for fire assay of many materials including techniques for ashing carbon and assays of Pt group metals.

How about further reducing the sample and take an aliquot of 5grms of pulverized sample and assay (have done this and it works so well with proper (SRMs).The idea being to start with whole lot of sample making sure you have a proper representative of the bulk. When ashing you limit your self to a smaller sample with doubting homogeneity because of how carbon behaves.

Thanks! Talking about variability of large samples does it mean that fire assaying directly will have a high variability of results to ashing before fire assaying?

Carbon by it's nature will be highly variable. Ashing simply reduces the reducing power of carbon so you can assay it as you would any sample. Sometimes you will need to ash the entire sample or a large number of small samples, as in a laboratory experiment, If you are taking samples from carbon adsorption columns run some triplicates. You should be careful ashing carbon to prevent contamination of the furnace. If you have a small furnace you can set aside for this function I would recommend you do that. If you choose to pulverize the sample and assay carbon directly be very careful. The carbon can easily contaminate your shatterboxes or pulverizers. Be sure to perform that task in an area especially designed for high grade materials. Run a test charge to determine the maximum amount of carbon you can add to your flux without your lead button exceeding 35 to 40 grams. If I recall correctly, that is the maximum reasonable size for a cupel to absorb. Have fun!

Depends on your goal. Accurate assays for metallurgical tests, you're better off taking a larger sample (we do 20g) and ashing first. Commercial and production labs will often assay only a small amount of carbon without pulverizing or ashing (~2g). Pulverizing leads to a high probability of contamination in the sample prep area and ashing requires time and equipment. Even a well calibrated XRF analysis on half a dozen granules may give you what you need.

So if you can accept a reasonable amount of variability (as in daily tracking numbers) you may be better off using a quicker, simpler method.

You are really making my day and your comment is giving me a lot of fun. But why one run triplicates on a sample from carbon adsorption columns.

XRF analysis is way expensive so I believe resorting to the old way of fire assaying is much economical.

Triplicates are used in umpire assays. It will also give you a feed for the variability of your sample. Active carbon columns will likely have less variability than you might think but running the triplicates will allow you to confirm that assumption.

Maya, your earlier comment on reducing litharge. That is the intent, to reduce to elemental Pb to collect PM's. The reducing power of the ore determines the amount of flour (C) required to obtain a manageable Pb button. Usually 1 or 2-g loaded C (as-is, w/o flour) in triplicate is used for assay. The assertions on ashing/contamination are correct. Fire assay crucible contamination must also be considered prior to reuse.

Hazen research has a lab that specializes in carbon based and coal projects. The methodology they employ for carbon analysis is to first take a larger sample, pulverize and blend. The blended solid is then placed in a LECO carbon analyzer. This is a method which has been used for the last 15 years and with proper inline filters has good repeat-ability.

You are correct. I should have been more specific in stating that excess carbon in the charge will result in a lead button that is either too large to cupel or results in an incomplete fusion. Adding flour as a source of carbon has been long practiced. I've done it myself. so you can direct fire a pulverized sample but you need to know exactly how much reducing power you are adding to the flux to achieve the proper size lead button.

There is good and bad with all the above methods. Let me say XRF a sample of the raw carbon first after grinding to fine powder. Then XRF a sample of loaded carbon. You will almost always have some value in the carbon. Now take 6, ten gram samples of the loaded carbon. grind them fine (250) mesh minimum. Take three and ash and follow up my firing adding enough carbon to achieve a 25 to 30 gram button. Take the other three and fire direct using 30 grams of litharge and slowly add about 3 grams of sodium nitrate, slowly. More if necessary to achieve a 25 to 30 gram button. Cupel all six buttons and compare the results with the XRF. If you want to achieve better results, flatten the lead buttons to paper thickness and electro-part. This can be done in sulfamic acid. In cupelling some values will oxidize and vaporize and some will go into the cupel.

Depends what your trying to achieve. For analyzing gold in a carbon sample. Pulverizing, ashing followed by direct aqua regia can give reliable gold assays and is relatively cheap compared to fire assay. Be-careful with you ashing sample size as too big a sample wont allow enough oxygen to ash the whole sample.

Unless you analyze carbon on a irregular basis, no lab will want to pulverize the carbon. especially if its fully charges + 2000g/t, and other lower grade sold and rock samples has to be tested in the lab and even if you have a dedicated room for carbon. I have found it very easy to reduce the possible contamination that may come from ashing by placing the carbon - approx. 2g in a crucible in which silica sand was previously placed. This form as seat for the carbon. after ashing and cooling, a thin layer of flux is spread over the ashed carbon to completely cover the ash. this is done prior to transporting the carbon to the next station ( fluxing). The entire content is transferred into a plastic bag in which a quantity of flux and additional borax or the pre calculated flux was already placed. The transfer is done by first placing the plastic bag snugly over the crucible mouth so that no air gets in or out and up turning up side down, the crucible with the sand, ashed carbon and the layer of flux. the bottom of the crucible is then tapped a few times to ensure that all material was poured out. I have found this method to be very efficient. each group of sample is assayed along with prepared in house carbon control. various grade of inhouse controls are prepared by soaking carbon in a prepared solution of cyanide matrix e.g 100ml of 100ppm cyanide matrix with 25g of virgin carbon. the solution is tested to ensure that the carbon has absorbed all the gold in the solution then dried and stored.