Printer Friendly

Treatment of Alaskan refractory gold ores: updating data for economic development of regional gold deposits.



The U.S. Bureau of Mines investigated treatment of gold and silver refractory ores and concentrates that respond poorly to conventional cyanidation techniques due to the complex mineralogy of the ore. Three types of ores were investigated: gold-locked placer-sands, gold locked in sulphides, and gold that is robbed from the pregnant cyanide solutions by minerals in the ore. Ore treatment included grinding, flotation, cyanidation, air oxidation, pressure oxidation, hypochlorite oxidation, roasting, and carbon in leach.

Worldwide investigations[1] have shown that possible reasons for poor response of gold-bearing material to cyanidation include:

* Physical lock-up (if the fine gold particles can be liberated by economic grinding the gold is locked, if they can not, the gold is encapsulated), * Carbonaceous materials (active carbon in the material adsorbs the gold from solution), * Decomposition of host minerals (sulphides decompose to form cyanicides and deplete the oxygen), * Coated gold (particles associated with sulphides in contact with a lixiviant can develop coatings), and * Insoluble alloys or compounds of gold (gold-bearing tellurides and arsenides, aurostibnite ([AuSb.sub.2]), and maldonite ([Au.sub.2] Bi)).

In addition, some sulphides, such as [Sb.sub.2] [S.sub.3] adsorb on gold particles as charged colloidal particles when in solution. Roasted lead-containing ores can contain lead compounds that coat the gold particles, rendering them insoluble to cyanidation.

For gold ores to react completely, certain requirements must be met. Gold must be in the form of discrete and clean particles, impurities that may inhibit the reaction must be absent, and an adequate supply of oxygen must be available.

A common technique to treat refractory ores has been roasting of flotation concentrates prior to cyanidation. Other techniques include fine grinding, pressure oxidation of sulphides, bacterial digestion, chemical pretreatment (preaeration and acid washing), and hypochlorite or chlorine oxidation.

As part of the overall objective of evaluating the potential of gold deposits in the Juneau mining district, test procedures were developed to identify the refractory nature of ores and to investigate methods for pretreating refractory ores prior to cyanidation. Four Alaska deposits that can be classified in one or more of the three refractory types mentioned above are Crow Creek, Haines Road Cut, Fern mine, and E Pluribus Unum.

Haines Road Cut and E Pluribus Unum deposits were investigated as part of the Juneau district study. The Fern and Crow Creek deposits were investigated as site-specific studies elsewhere in Alaska (Anchorage and Willow Creek mining districts). Cyanide amenability tests determined that these ores were refractory in nature. Tests on these refractory ores included air oxidation, hypochlorite oxidation, acid pressure oxidation, roasting, and carbon in leach.

Gold recovery from gold locked in sulphide ores, represented by the Haines Road Cut, was improved by proper grind size and by preoxidation. Recovery increased from 65.3% by cyanide leach to greater than 83.6% by pressure oxidation-leach of a flotation concentrate.

Gold loss by absorption on gold robbing material was reduced by oxidation. The Fern ore required minimum oxidation treatment to decrease the gold loss from the preg solution, while the E Pluribus Unum required pressure oxidation treatment to obtain similar reduced gold loss.

The recovery from gold-locked placer sands, as represented by the Crow Creek deposit was improved by grinding. Gold recovery increased from 41.2 to 97.7%.


The Crow Creek gold placer[3,4] is located about 30 miles east of Anchorage. This material is a refractory gold ore because only 41% of the jig concentrated gold (28 oz/st Au and 16 oz/st Ag) was recoverable by amalgamation. Mineralogical examination showed that about one-half of the sample comprise mica schist. The remainder was largely a mixture of iron oxides (magnetite, hematite, and limonite). Iron pyrite made up several percent of the sample. Other commonly occurring minerals were scheelite, galena, zircon, and native gold. The native gold was abundant. Some of the gold contained inclusions of iron oxide and about 20% of the gold reports with the magnetic fraction of the sample as shown in table 1.

The Haines Road Cut was discovered by USBM personnel in 1986, and is located three miles south of Haines. Chemical analysis of the ore showed it contained up to 0.33 oz/st Au, 0.88 oz/st Ag, and 8% Cu. This is the first significant report of such mineralization in the Chilkat volcanics and opens up a large area to exploration for similar mineralization. A mineralogical examination of the samples of the ore showed that the major minerals present were chalcopyrite, quartz, and pyrite. Smaller amounts of malachite, chromite, magnetite, ilmenite, pyroxene, K-feldspar, sphalerite, covelite, bornite, rutile, chalcocite, and arsenopyrite were present in the sample. Traces of lead, nickel, and scandium were also detected.

Several gold metallic grains were picked from a gravity concentrate and examined by scanning electron microscopy (SEM). The grains contained about 80% Au, 15% Ag, and the remainder as iron and chromium minerals.

The Fern Mine[5] is located 50 miles northeast of Anchorage. Ore from the mine was processed in the past by flotation and then treated by amalgamation to recover the gold. Analysis of the ore sample showed as much as 7 oz/st Au and 0.19 oz/st Ag. Samples that represent the deposit consist of gold-bearing quartz vein material. The veins contain ankerite, calcite, and considerable clay, probably the result of hydrothermal alteration. The veins are hosted by diorite and locally contain scheelite. Concentrates from bulk sulphide flotation contained visible free gold (30 oz/st Au) with much of it being 24 mesh (0.707 mm), plus pyrite, chalcopyrite, and perhaps other sulphide minerals. In addition, the ore contains some carbonaceous material that was determined to be a preg robber; that is, the carbonaceous material is active so that when it comes in contact with gold in solution it adsorbs the gold from solution. The apparent gold extraction is lowered by the amount of gold adsorbed.

The E Pluribus Unum[6] was discovered before 1904 and developed as an underground mine in southeast Alaska about 24 miles northwest of Juneau. Samples from the deposit contained about 0.24 oz/st Au. Mineralization consists of three types of material: (1) a low-grade stringer lode, (2) medium-grade massive quartz veins, and (3) high-grade lenticular pods. The stringer lode and the quartz veins have pyrite and arsenopyrite. The pods contain arsenopyrite, galena, sphalerite, stibnite, and 4.8 to 14.5 oz/st Au. Carbonaceous material was also found in the samples.


To evaluate the ores, a testing routine was developed that included mineralogical examination with a SEM on a minus-10-mesh sample to determine the minerals and their liberation size, and cyanide amenability tests leaching ground ore in a rolling bottle.

The cyanide amenability results shown in table 2 give a low extraction for the Haines Road Cut, 65%. This fact, combined with the determination of substantial sulphides by SEM analysis suggest refractory tendencies. The sulphides may be depleting the oxygen and/or the cyanide, and/or the gold is locked in the sulphide matrix.

The extraction percentages for the Fern and E Pluribus Unum ores were not particularly low but the residue grade is still high and suggests further treatment may improve the extraction. Additional evidence of their refractory behavior is seen in the target grind test results discussed below.

The Crow Creek sample was treated by amalgamation because of the high concentration of gold in the sample. Amalgamation is more convenient than cyanidation for recovery of gold from concentrated samples. The Bureau research was performed using amalgamation to provide meaningful information. The results of the amalgamation test indicate that the gold is not free and is difficult to extract, and since this is a placer-sand-deposit lockup was suspected.

The target grind size test follows the cyanide amenability test in the usual evaluation of any gold ore. The purpose is to find the size where an acceptable tail grade will be obtained. In addition, refractory tendencies can be seen from the results. Target grind size tests were conducted by leaching samples that had been crushed to minus 0.5-in. ore, the pH adjusted to 10.5 with lime and 20 lb/st NaCN. The slurry was agitated in a rolling bottle for 72 hr. After leaching, the sample was filtered and washed. The leach residue was screened into size fractions and each size fraction assayed. From the results of this test the required grinding for gold and silver extraction was estimated. Generally a decrease in the grind size shows a decrease in residual gold in the residue. However, in the case of some refractory gold ores this is not true. The results of target grind tests for Fern and E Pluribus Unum ores are given in table 3, they show that for the larger size fractions the gold recovery with cyanide leaching increased with decreasing mesh size as is normally expected. However, at the finer size fractions the residue grade begins to rise until at minus-325-mesh gold in the cyanide leach residue increased significantly, indicating particle locking or gold robbing.

Gold extraction from the Fern ore was 87% with cyanide and 61% for the E Pluribus Unum.


Gold ores containing sulphides can be refractory to cyanidation. These ores typically contain sulphides[2] that include pyrite ([FeS.sub.2]), pyrrhotite ([Fe.sub.5] [S.sub.6]), arsenopyrite ([FeS.sub.2].[FeAs.sub.2]), realgar (AsS), orpiment ([As.sub.2] [S.sub.3]), stibnite ([Sb.sub.2] [S.sub.3]), galena (PbS), anglesite ([PbSO.sub.4]), and many of the copper minerals. The sulphide minerals can react with oxygen in the cyanide solutions, thus depleting the solution of oxygen, and others not only consume oxygen but form cyanicides, depleting the cyanide content.

The Haines Road Cut is an example of a sulphide containing gold ore that is refractory to cyanidation. Direct cyanide leaching of this ore yielded 64% of the gold being extracted. To improve gold recovery the ore must be further treated. Treatments included air oxidation and acid pressure oxidation of the concentrate followed by cyanidation.

Air oxidation tests were initiated by ball-mill grinding to greater than 65% minus-325-mesh at 50% solids. The slurry was then transferred to a bottle, the pH raised to between 11 and 12 with lime, and the sample agitated and air sparged for 24 hours. Cyanidation of the resulting slurry increased gold extraction to 73%.

Acid pressure oxidation tests were conducted on a bulk flotation concentrate containing 86% of the gold [concentrate produced with Cyanimide A-208 (sodium diethyl and sodium di-secondary butyl dithio-phosphate), potassium amyl xanthate, copper sulfate and frother] at 20% solids, 220 [degrees] C, and with 50 lb/in.02 [O.sub.2] for two hours. Cyanidation of the leach residue resulted in greater than 96% gold extraction; total gold extraction was greater than 83.6%.

A summary of the air oxidation and acid pressure leach tests followed by cyanidation is given in table 4.

Improved overall extraction for the pressure oxidation treatment will result if the flotation is more efficient. Finer grinding can be beneficial to a point and then counter productive. Splits of the sample were ball-mill ground so that 62, 80, and 90% of the each split passed 325 mesh. The gold was concentrated in each split by bulk sulphide flotation. The results of these tests are given in table 5. This table shows the extraction drop as the slimes begin interfering with the flotation.


Ores containing minerals that cause gold to redeposit from the leach solution are considered gold robbers. Carbonaceous or carbon-bearing materials usually cause this problem and must be deactivated by oxidation or by some other approach. Even though 80% of the gold is recovered during cyanide amenability and testing is continued, it is important to beware of gold robbing indications. The E Pluribus Unum and the Fern ores exhibit this behavior.

The purpose of the gold rob test is to determine if the ores suspected of being a gold robber will adsorb gold from the cyanide leach solution. A gold-bearing solution, which was prepared by dissolving [AuCl.sub.3] in a cyanide solution and then neutralizing the excess cyanide, was contacted with a ground sample of each ore. The [AuCl.sub.3] solution was made basic with sodium hydroxide prior to cyanide leaching. Excess cyanide in the gold robbing solution was destroyed to prevent the leaching of any gold from ore and thus biasing the test results. Results are given in ounces of gold that is robbed from the known solution per short ton, a decrease in the gold that is robbed shows the degree that the gold robbing material is neutralized. Test showed that the Fern and the E Pluribus Unum mines contained gold robbers.

The gold robbing material in the Fern Mine and the E Pluribus Unum ores is carbonaceous matter. Early operations at the Fern mine recovered the gold by flotation and amalgamation and thus avoided the gold robbing problems.

A gold robbing ore can be treated either by neutralizing the robbing portion by oxidation or by competing with the robbing portion during the leaching of the gold by adding carbon to the leach slurry. Tests conducted to reduce the gold robbing effects of the ore included preoxidation with air, preoxidation with hypochlorite, acid pressure oxidation, roasting, and carbon-in-leach (CIL).

The results of the gold rob treatment are listed in tables 6 and 7. They show that the Fern ore required only minimal treatment to prevent robbing of the gold, while the E Pluribus Unum required severe treatment such as acid pressure oxidation or roasting. The gold robbed by the Fern decreased to 0.06 oz/st with air oxidation; whereas, a pressure oxidation leach was required for the E Pluribus Unum to achieve similar results, 0.02 oz/st.

Table 7 compares results of CIL tests with cyanide amenability tests. The data show CIL improves gold extraction from the Fern Mine sample from 81.7 to 89.3%. The CIL did not improve gold extraction from the E Pluribus Unum mine sample. The gold extraction was 80.6 and 79.8%, respectively, for cyanide amenability and CIL tests.


In the Crow Creek ore the gold is locked in a placer sand. This is an example of a material for which particle size reduction is all that is needed to improve gold recovery. Typically, a placer sand is processed by gravity separation techniques. A sample is screen sorted into size fractions and then the size fractions that contain the gold are further treated by a variety of gravity techniques. This particular sample of Crow Creek placer sand was concentrated with a jig. The jig concentrates from the property contained from 2 to 400 oz/st Au. By the nature of the gold, individual splits show a wide fluctuation in assay because the coarse particle size of the gold made it difficult to assay a representative sample. The initial test on the first Crow Creek sample showed that grinding to 67% minus-325-mesh increased the gold recovery from 41.2 to 97.7%. The test results are shown in table 8.

The principle of size reduction to improve recovery was also demonstrated with a lower grade sample (1 to 2 oz/st Au and about 1 oz/st Ag), as shown by the test results in table 9. Splits of the sample were ball mill ground so that 67, 78, and 89% of the each split passed 325 mesh. The samples were amalgamated to recover the gold. The results show that the amalgam tail decreased from 0.29 to 0.044 oz/st when the ore mesh size was decreased from 67 to 89% minus-325-mesh, respectively. There was poor agreement between head sample assays because of the occurrence of abundant free gold in the samples. [Tabular Data 1 to 9 Omitted]


[1]Dry, M. J., and F. B. Coetzee. The Recovery of Gold From Refractory Ores. Gold 100. Proceedings of the International Conference on Gold. V. 2: Extractive Metallurgy of Gold. Johannesburg, SAIMM, 1986, pp. 259-274. [2]Nagy, I., P. Mrkusic, and H. W. McCulloch. Chemical Treatment of Refractory Gold Ores - Literature Survey. Natl. Inst. for Metall., Rep. 38, 1966, pp. 3-25. [3]Hoekzema, R. B., and S. A. Fechner. Placer Gold Sampling In and Near the Chugach National Forest, Alaska. BuMines IC 9091, 1986, 42 pp. [4]Hoekzema, R. B., S. A. Fechner, and J. M. Kurtak. Evaluation of Selected Lode Gold Deposits in Chugach National Forest, Alaska. BuMines IC 9113, 1987, 56 pp. [5]Kurtak, J. Results of 1984 Bureau of Mines Site Specific Field Studies Within the Willow Creek Mining District, Alaska. BuMines OFR 17-86, 1986, 25 pp.; NTIS PB 86-163904. [6]Redman, E., W. S. Roberts, A. Clough, and J. Kurtak. Preliminary Mine, Prospect, and Sample Location Maps and Descriptions, Juneau Gold Belt Area. BuMines OFR 85-86, 1986, 68 pp. [7]The Australian Institute of Mining and Metallurgy. Carbon-in-pulp Technology for the Extraction of Gold. Clunies Ross House, 191 Royal Parade, Parkville; Victoria, Australia, 3052, 1982, 442 pp. [8]F. M. Hamilton, B. A. (oxon). Manual of Cyanidation, McGraw-Hill (New York), 1st ed., 1920, pp. 173-187.

PHOTO : Fig. 1, Ore from the Fern Mine--the portal is situated near this valley--was previously

PHOTO : recovered by flotation and amalgamation methods.

PHOTO : Fig. 2, The south-end sites, which were sampled for gold in sulphide ores, are located in

PHOTO : the Haines Road cut.

PHOTO : Fig. 3, Gold-locked placer sands in the study originated from the Crow Creek Mine, which

PHOTO : is immediately above the small valley.

PHOTO : Fig. 4, The ore from the E Pluribus Unum's lower adit required pressure oxidation to

PHOTO : reduce the gold loss. The inset photograph shows iron oxide attached to gold from the

PHOTO : Crow Creek mine.

W. R. McDonald, metallurgist J. L. Johnson, chemical engineer; and R. G. Sandberg, research supervisor U.S. Bureau of Mines
COPYRIGHT 1990 PRIMEDIA Business Magazines & Media Inc. All rights reserved.
No portion of this article can be reproduced without the express written permission from the copyright holder.
Copyright 1990 Gale, Cengage Learning. All rights reserved.

Article Details
Printer friendly Cite/link Email Feedback
Author:McDonald, W.R.; Johnson, J.L.; Sandberg, R.G.
Publication:E&MJ - Engineering & Mining Journal
Date:Jun 1, 1990
Previous Article:Heap leaching for precious metals.
Next Article:Barrick granted partial judgement in Mercur lawsuit.

Terms of use | Privacy policy | Copyright © 2019 Farlex, Inc. | Feedback | For webmasters