The method of treating carnotite, a uranium, ore used by the Bureau of Mines is outlined in this chapter.

Leaching, Filtering, and Washing Ore

The ore is ground to 20 mesh and is leached with strong hot nitric acid in acid-proof earthenware pots. The amount of acid used is 121 pounds of 100 per cent nitric acid to 500 pounds of ore, the acid being diluted to 38 per cent strength. However, the strength may be varied somewhat, ores high in sulphates requiring stronger acid. The acid is brought near the boiling point by live steam, which has been passed through a baffle in order to eliminate any impurities from the steam. The ore is then run in, being stirred with a wooden paddle during the process. The acid is heated for 15 minutes longer, with occasional stirring, and the acid is then run into an earthenware vacuum filter, asbestos filter cloths being used. As much as possible of the sand is held back in the pot by means of a long wooden plug manipulated by hand, and this sand is given an acid wash with acid about one-third the strength of that used for the first leaching. This sand is then dumped on the filter, and receives two washings with hot distilled water.


The whole work from start to finish, leaching, filtering, and washing, can usually be done in about seven hours. The residue is thrown on the dump and the filtrate is run through earthenware pipes into a large precipitating tank made of California redwood, where it is diluted by the addition of water. This solution is stirred, and sodium hydroxide is run in slowly, with the object of reaching as nearly as possible the neutral point without forming a permanent precipitate. If too much alkali is added, both iron and vanadium are precipitated, and contaminate the radium-barium sulphate; on the other hand, if not enough sodium hydroxide is added, the acidity remains too high and the solvent action of the nitric acid on the radium-barium sulphate is not sufficiently decreased. A little practice enables the operator to get the right point by visual observation without titrating, titration not being possible because of the large amount of dissolved material that would be precipitated by an alkali.

A solution of barium chloride is then added, usually in the proportion of about 2 pounds of barium chloride to 1 ton of ore, and after the liquid has been thoroughly stirred, sulphuric acid is slowly added, as stirring continues. Fifteen pounds of sulphuric acid to 1 ton of ore is the preferable quantity. The stirring is continued for one hour, when the whole solution, containing the barium-sulphate precipitate, is elevated to a conical settling tank by means of a centrifugal pump having parts that come in contact with the liquid made of duriron. The solution is then allowed to settle in the conical tank for a few days. In the original plant this period was three days; in the second plant it was increased to four. By means of a floating siphon, the clear solution is decanted into a tank containing an excess of boiling sodium carbonate, where the iron, calcium, and most of the aluminium are precipitated, and the uranium and vanadium go into solution as the double carbonate of uranium and sodium, and as sodium vanadate. The solution must be boiled for at least three hours after all the acid solution has been run in, as otherwise there is considerable loss both in uranium and vanadium, which remain with the iron precipitate.

PART 1: Radium, Uranium & Vanadium Extraction & Recovery from Carnotite


PART 3: VANADIUM & URANIUM Extraction and Recovery

PART 4: RADIUM Extraction & Recovery

PART 5: Processing, Extraction & Recovery of RADIUM 

The radium-barium sulphates and the associated liquor are run onto an earthenware suction filter, filtered, washed, and finally treated with a dilute solution of sodium hydroxide in order to remove the last traces of free acid. The filtrate is run into the carbonate tank with the rest of the acid liquor. The radium-barium sulphates are placed in iron pans and dried in a hot-air oven.

The process of Radium, Uranium & Vanadium Ore Processing is clearly indicated in the diagram following:diagram-of-steps-in-bureau-of-mines-method-of-radium-extraction

Treatment of Sodium Carbonate Solution

The sodium carbonate solution, carrying the uranium and vanadium, is nearly neutralized with nitric acid, the solution being constantly stirred by means of compressed air; then sodium hydroxide is added to the boiling solution until there is a complete precipitation of sodium uranate. No attempt has been made to control definitely the color of this precipitate, as the sodium uranate has always been converted into oxide for final sale. In addition, the sodium uranate has always carried vanadium, as precipitation of the uranate in the presence of vanadium has so far always caused the precipitate to contain at least 7 or 8 per cent of V2O5. Consequently, it is generally necessary to remove the vanadium from the sodium uranate before it can be sold as such. Some early experimental work showed that redissolving with sulphuric acid and reprecipitating with sodium hydroxide would be necessary three or four times before the vanadium could be reduced to less than 1 per cent. This method of refining, therefore, could not be used commercially, and it became necessary to find some other cheaper and more efficient process. Such a process is described in subsequent pages.

Treatment of Vanadium Solution

The hot solution from the sodium uranate is completely neutralized with nitric acid, air being blown into the liquid in order to eliminate the carbon dioxide. Ferrous sulphate is then added, the liquid being continually agitated, and the precipitate of iron vanadate is filtered and washed.

The grade of the precipitate depends largely on the final acidity of the solution. If the solution is slightly acid, a high-grade precipitate carrying 40 or 42 per cent V2O5 may be obtained. On the other hand, some of the vanadium is not precipitated and is therefore lost. If the solution is perfectly neutral after the addition of the ferrous sulphate, a lower grade product is obtained, carrying 30 to 33 per cent V2O5, but all of the vanadium is precipitated. It is advisable not to boil the solution after the addition of the ferrous sulphate, although the solution should be hot at the time of this addition.

Treatment of Sodium Nitrate Solution

The filtrate from the iron vanadate is almost wholly a solution of sodium nitrate, the main impurity being a moderate amount of sodium sulphate. The solution is evaporated in iron tanks heated by steam under pressure. Air from a compressor is blown into the solution and evaporation is rapid. As soon as the solution is concentrated enough it is run into steel crystallizing pans where it crystallizes. After draining on draining boards the crystals are collected without further treatment and employed to make fresh nitric acid for use in the plant. As the losses of sodium nitrate are not great, the actual results have been to reduce the cost of the nitric acid below the purchase price of hydrochloric acid of the same acidity.

Recoveries Made by Bureau’s Process

The extraction and recovery of the radium have been excellent even from the start, an extraction of more than 90 per cent having been attained with many carloads of ore. The extraction of the uranium is also practically complete, but a considerable part of the vanadium is left in the ore. In fact, the presence of much vanadium in the ore is a disadvantage, as in the leaching pot the vanadium has a tendency to separate out as vanadic acid, which retards filtration considerably, and tends to reduce the extraction of the radium. Recovery of a few per cent more of the radium content more than compensates for a loss of 50 per cent of the vanadium, and if both can not be recovered, preference must be given the radium. The vanadium in the carnotite goes into solution readily, but roscoelite and other vanadium minerals present are decomposed with much difficulty, and it is not possible on a commercial scale to extract all of the vanadium in the ore, although complete extraction can be attained in the laboratory if enough acid is used.

General Description of Denver Plant and Equipment

When the plant of the National Radium Institute was projected, it was to be entirely experimental. At the same time, plans were made to build and equip it so that if the experimental work were successful, operation on a commercial basis would be possible by enlarging the plant or by using it as built. Therefore, it was necessary to design, erect, and equip a building that could be used permanently on a commercial scale, and yet to keep the cost down so that if the experimental work were a failure the loss would be as low as possible. For this reason, some equipment was not put into the plant that would have made the original work a little more efficient if the initial investment had been a little larger.

After the experimental work had proved successful, the officials of the National Radium Institute decided that they wished the radium delivered more rapidly than was possible with the original small plant (Pl. I, A). Consequently, an additional plant (Pl. I, B) was built adjoining the first one. Both were equipped as a separate unit so that they could be run separately or at the same time. For convenience of description, these plants are mentioned in this report as the “first” plant and the “second” plant.

Refining of the radium concentrate is done in a separate building, usually termed the “sulphate building.”

During the first six months of work, nitric acid was purchased, but the sodium nitrate was recovered and stored. In December, 1914, and January, 1915, a nitric acid plant was erected, and since then the institute has made its own nitric acid, from the recovered sodium nitrate.

All the buildings of the plant proper are of frame construction, with an outer covering of galvanized iron painted on the inside with one coat of graphite paint to protect it from acid fumes. The roofs of the buildings are of elaterite.

The original or “first” plant (Pl. I, A) is 80 feet long, 45 feet wide, 21 feet high on the south, sloping to 18 feet on the north. Adjoining this is a storeroom (shown at extreme left of Plate I, A) with doors opening on to a switch that runs past the plant, connecting with both the Denver & Rio Grande and the Colorado & Southern Railroads. The original storeroom was only one story high, but later an additional story was added, as well as a small grinding and sampling plant, which was placed in a room built into one corner of the storeroom.

The “second” plant is 130 feet long, 30 feet wide, and 24 feet high on the north, sloping to 21 feet on the south. The old boiler room was turned into a furnace room for the refining of the uranium, and for the preliminary treatment of the radium-barium sulphates. A new boiler room to serve both plants was erected at the same time as the second plant. It is shown at the right of Plate I, B; a plan and sections of this budding are shown in Plate II.



Description of First Plant

The floor plan of the original or “first” plant is shown in Plate III and the plan of the upper landings on the south side in Plate IV, A. The ore and acid are carried to the upper landing by means of the elevator (see Pl. III), the acid being transported in carboys. The handling of the acid in the carboys has been fairly satisfactory, but an improvement could be made by conveying the acid in glass pipe lines to the leaching pots (27a, 27b, etc., Pl. IV, A), which would partly eliminate the handling of the acid by hand. This was one of the changes that it was not thought necessary to make under the conditions.

Construction and Operation of Leaching Pots

Each of the earthenware pots in which the leaching is done has a capacity of 107 gallons. Originally each was set in an iron container holding oil that was heated with a steam coil. It was thought that some external method of heating would be necessary in order to quickly raise the temperature of the acid to the desired temperature and to maintain it at this temperature during the period of heating. The results were not satisfactory, as many of the leaching vessels cracked, probably from somewhat uneven heating and the strains to which they were subjected by the packing. During the early stages of the work, in addition to the heating by the oil bath, live steam was run into the acid through glass tubes connected with a pipe leading from a large baffle (20, Pls. III and V) that served to eliminate, as much as possible, impurities from the steam. Experience showed that the acid could be heated rapidly enough with live steam alone, and therefore the oil bath was discarded and sand was substituted. The results obtained were an improvement, not so many of the leaching vessels breaking, and the acid being heated at a satisfactory rate. With a view to getting still better results, cement jackets were finally tried, and two or three experimental pots were set in concrete. The results were so good that sand packing has been discarded. The breakage has been small and the cement jackets have proved satisfactory in every way.

The leaching pots are partly covered with wooden covers, each of which has a wooden flue leading to a main flue that goes through the roof (see Pl. IX, B, p. 46). A small steam pipe is placed in the main flue so that a jet of steam can be admitted when desired. By this means, practically all of the fumes from the acid can be removed without any trouble. The wooden covers of the pots and the wooden flues are coated with graphite paint, which makes a good protective agent. Stirring is done by hand with wooden paddles (Pl IX, B), as it is almost impossible to install mechanical stirrers that would be satisfactory under the conditions.

During the period of leaching the spouts of the leaching pots are closed with rubber stoppers fastened to cast-iron plates attached by sleeves and set screws to horizontal rods, which are turned by levers inserted through the floor of the upper landing (fig. 1). The stoppers can be removed from the spouts or replaced by means of these levers, which are worked from the upper landing, so that all danger from the splashing of hot acid during the dumping of a pot is eliminated. Such splashing, however, can be reduced to a minimum by inserting a wooden plug at the end of a handle in the upper part of the outlet of the pot before the stopper is removed. When the stopper is taken out and the acid begins to run from the pot, the plug is sucked into the opening. Raising the plug when necessary permits the acid to flow onto the suction filter below at any desired rate.

The acid wash is contained in a small wooden tank (36, Pl IV, B). From this tank it runs by gravity through an ordinary acid-proof rubber hose to the leaching pots for washing the ore.

Construction and Operation of Suction Filters

Plate IV, B, showing a section of the first plant, indicates the arrangement of the leaching pots (27a, 27b, etc.) and of the suction filters (28a, 28b, etc.) below. These filters were made by the German-American Stoneware Works Co., and are of the “Edda” type (see Pl. IX, C, and fig. 1). The upper part of each filter has a capacity of 105½ gallons, as has the lower part. Each bottom vent has a ground-in bibcock, and the upper opening is connected by means of about 6 feet of ½-inch pressure rubber tubing with a pipe running to the vacuum pump. This pipe first connects with a baffle (47, Pls. VI and VII), which protects the pump itself. This baffle is filled with vitrified brick, and a strong solution of sodium hydroxide constantly circulates through it.

Each filter is set on a small truck (see Pl. IX, C) running on rails so that when the residue is to be removed from the filter the truck can be pushed or pulled from, underneath the leaching vessels. The filtering medium consists of an asbestos filter cloth covered with 1 inch of coarse sand over which there are several strips of wood. These strips are held down by three or four bricks suitably placed.

The suction filters have lasted even better than the leaching pots. Before the hot acid is dumped, a small piece of steam hose connected with a steam line is pushed up through the bibcock of each filter so that the lower part is slowly steamed and the temperature raised gradually. Another similar piece of hose, connecting with a second steam fine, is pushed down into the upper part of each filter and the

plan and sections of boiler

section showing leaching pot

whole covered by any suitable material as oilcloth or sacking. In this manner, the upper part of the filter is also steamed and the temperature gradually raised before the hot acid is dumped. With these precautions, few breaks have occurred. The acid from the suction filters flows through a horizontal, acid-proof earthenware line which dips slightly toward the radium precipitating tank, and has an opening below the bibcock of each suction filter.

Precipitating Tanks

The precipitating tanks (Pls. III and VII) are all of 2-inch California redwood, which has satisfactorily withstood the action of acids and of weak alkalies. The radium precipitating tank in the first plant, at the time of writing (September, 1915) has been in use 15 months, and is still in fair condition, although it has been subjected daily to the action of warm dilute nitric acid. Those tanks in which solutions are boiled have a wooden cover, with a wooden flue 12 inches square that extends up through the roof, so that little steam escapes within the plant.

The bottom part of each settling tank is conical, although the vertical staves, which constitute the outside of the tanks, extend the entire length. In other words, these tanks are ordinary 8½ by 5 redwood tanks with a cone inserted in the bottom, and can be supported from below instead of being suspended.

There was considerable difficulty in getting these tanks tight. Ordinarily, wooden tanks are built to hold water and any small leaks are closed by the swelling of the wood. When acid solutions are poured into wooden tanks, the wood tends to contract, rather than to swell, so that acid tanks must be set up differently. The staves must be properly machined and all joints made with special care, extra lugs must be used, and the use between the staves of a rubber cement is advisable. A tank must be absolutely tight when liquid is first put in. However, if these precautions are taken, a tank can be kept tight without serious trouble, an occasional tightening of the lugs or a little calking being all that is required. Oval holes were cut in the lower part of the tanks large enough to allow a workman to crawl in and thus get at the lugs holding the conical part of the tank (see Pl. X, A, p. 50). The lugs and, indeed, the whole exterior of the tanks should be kept covered with graphite paint. The other tanks were put up with a thin coating of white lead between the staves.

It has been found convenient to run the sodium carbonate into the tanks by bringing it on a small truck by means of the elevator (Pl IV, A) to the third landing, and then letting it fall into the tank below through a wooden chute 8 inches square.

floor plan of old plant

plan of upper landing

section of first plant

Steam-Heating Devices

Each tank in which solutions are boiled contains a grid composed of 1½-inch Byers pipe with cast-iron fittings, the extent of heating surface in the different tanks depending, of course, on the object to be attained. In the larger tanks (10 to 12 feet in diameter) in which liquids are heated without being evaporated, 12 lengths 6 feet long have been found sufficient. In the nitrate storage tanks, where a certain amount of evaporation is desired, the heating surface is nearly twice that for the larger tanks. The steam flowing to these grids passes through a large baffle (20, Pls. V, VI, and VII) filled with vitrified brick, which eliminate impurities carried from the boiler. As the grids are also connected with suitable traps, a constant supply of distilled water is obtained and stored in a large iron tank (8, Pls. V, VI, and VII). From this tank distilled water is piped to the third landing to a point near the leaching pots, and by means of compressed air the water is elevated to this point whenever desired. As it is usually hot considerably less steam is needed in connection with leaching than if the water were cold. Exhaust steam is used for preliminary evaporation of the nitrate solutions.

Filter Presses

The filter presses used in the first plant are of the Shriver two-eyed “washing” type. The two larger presses (33a, 33b, Pl. IV) contain 30 plates, each 24 by 24 inches. The smaller presses (33c, Pl. IV, and 33d, Pl. V) contain 21 plates each, 18 by 18 inches. The iron, precipitate press (33a, Pl. IV) is filled twice during the day’s run, so that if one desired to empty this press only once a day, its capacity would have to be doubled. The small presses take care of the uranium and vanadium at one operation. Besides being connected with the tanks, the presses are piped for both water and compressed air. The pumps (16 to 19, Pl. III) connected with these presses are of the brass-lined, double-acting Worthington type.

Acid Conveyors

The acid from the radium precipitating tank is elevated to the settling tanks by means of a centrifugal pump through 2-inch acid rubber hose. The parts of the pump that come in contact with the liquid are made of duriron, and have resisted the acid excellently, as has the rubber hose, which is lined with pure gum. The acid is transferred to any desired settling tank by running it through the hose into a wooden flume connecting the different tanks.

Floating Siphon

The floating siphon (Pl. X, B, p. 50) consists of a square wooden frame through the center of which a 1¼-inch acid hose extends about 18 inches, or any other desired length. The frame is built so that it settles in the tank until at a certain point it rests on the sloping sides of the cone. The length of hose going through the frame is long enough to siphon off the clear liquid above but not to disturb the radium-barium sulphate precipitate. The other end of the hose is connected with a 1½-inch piece of glass tubing 5 feet long, from which another piece of the same type of hose runs into the tank. An earthenware stopcock is inserted at a convenient point near the glass tube, and is held by a chock so that the cock can be turned but can not be forced out by the pressure of the liquid. In this way the flow of the liquid can easily be controlled, and at the same time the liquid can be examined as it runs through the glass tube. As long as it is not turbid it is free from radium-barium sulphate precipitate. If the workmen, in occasional inspection of the tube, note any turbidity they can shut off the siphon and thus eliminate radium losses. In running the acid liquid into the carbonate tank, any convenient form of spreader can be used to prevent much acid striking the carbonate solution at any one point and thus reducing too much the alkalinity of the liquid at that point.

Miscellaneous Features

After the clear liquid has been siphoned off, the radium-barium sulphate with the remaining liquor is introduced onto earthenware suction filters, which are placed below the settling tanks. As mentioned previously, these filters are also of the Edda type, the upper and lower parts each holding 52¾ gallons of liquid. They are placed on small trucks running on tracks, so that they can be pulled from underneath the settling tanks and the radium-barium sulphate easily removed (Pl. X, A).

A compressor delivering about 300 cubic feet per minute was originally situated on the north side of the plant (21, Pl. III). When the second plant was built, however, this compressor was removed and a larger one (55, Pl. II) was placed in the boiler house and now serves both plants.

The mechanical stirrers in the different tanks are belt driven, one motor driving the stirrers in tanks 1 and 2 (Pl. III, and Pl. IV, B), and another one driving those in tanks 3 and 4. The liquids in tanks 5, 6, and 7 are agitated by compressed air, as the weight of the precipitates is not large and the time of agitation is not long. The stirrers in tanks 1 and 2 are of wood and have two blades revolving 8 inches and 18 inches above the bottom of the tank, and the stirrers

section reversed of first plant

plan of new plant

in tanks 3 and 4 are of iron, and each has one blade revolving just below the heating coil and a second blade about 6 inches above the heating coil.

Description of Second Plant

The general arrangement of the second plant is similar to that of the first, but as the building had to be of a different shape, some changes were made, as well as some improvements.

Instead of using Shriver presses, three 25-inch, lever-operated, clam-shell Sweetland filter presses (Pl. VIII) were installed. Each one of these has 150 square feet of filtering surface. One takes care of the iron precipitate from the sodium carbonate tank, the other handles the uranium, and the third the vanadium. As the total capacity of the last two presses is larger than is required by the one plant, the uranium and vanadium from both plants are now filtered through these presses, the solutions carrying the precipitated uranium and vanadium in the old plant being pumped into the corresponding tanks in the new. These presses are piped for water, compressed air (on both sides of the cloths), vacuum, and steam. It has been found difficult to get the cakes from the presses sufficiently dry to be dumped easily without the use of compressed air.

The filtrates from the iron and uranium precipitates go directly from the pumps into elevated tanks (51, Pl. VII). From these the iron filtrates are run by gravity into the uranium-precipitating tank (6, Pls. III, VI, and VII), and the uranium filtrates into the vanadium tanks (7, Pls. III, VI, and VII, and 39, Pl. V).

The stirrers in the new plant are driven by worm gears instead of belts. The stirrer in the radium-precipitating tank under load makes 44 revolutions per minute and that in the sodium carbonate tank makes 16 revolutions per minute. Each stirrer is driven by a 2-horsepower motor. The duriron centrifugal pump used for elevating the liquid from the radium-precipitating tank is the same size as was used in the first plant; it is rim at 2,000 revolutions per minute and is belt-driven from a 5-horsepower motor.

The sodium hydroxide, which is used for partial neutralization of the acid in the radium precipitating tank, and also for precipitating the uranium, is made up in a steel tank (49, Pl. VII), placed on an exterior platform connecting the new plant and the boiler house. The sodium hydroxide in the iron drum is broken with a sledge hammer, and dumped into this tank, the solution being agitated with compressed air. It is advisable to use comparatively little hot water until all of the sodium hydroxide has been dissolved and the solution has cooled somewhat, when it can be diluted. The solution is then run by gravity through a pipe into a large steel storage tank (45, Pls. VI and VII), from which it can be transferred by means of compressed air to the measuring tanks (30, Pls. IV and VII), situated on the upper landing in each plant. By means of a float attached to a cord terminating in a weight, which rises and falls over a graduated scale, the amount of liquid taken from these tanks can be determined. From each of these measuring tanks pipes run to the radium precipitating tank and to the uranium tank in each plant.

The system for obtaining distilled water for the new plant is the same as that in the old, the distilled water reservoir (8, Pls. VI and VII) being correspondingly larger.

Sulphate Building and Equipment

The small building in which the radium barium sulphates receive preliminary treatment and fractionation (figs. 2 and 3) is situated

plan of sulphate building

just behind the first plant. The arrangement as shown in the plan (fig. 2) is not exactly the arrangement that actually existed at the time of writing (September, 1915) because the silica-lined, acid-proof kettles ordered in France had not arrived; the arrangement used was more or less tentative. The plan shows the arrangement as it will be when completed.

section of second plants

In order to crystallize radium-barium salts from acid liquors, it is, of course, necessary to a certain extent to evaporate the solutions, in order to get the concentration necessary for proper crystallization.

section of sulphate building

This might be done in earthenware vessels, provided a satisfactory and efficient method of heating could be obtained. Copper or silver steam coils might be used, but both of these metals are gradually attacked by concentrated hydrochloric acid, and in addition there would be a tendency for crystals to form around the coils. Breaking these crystals away would tend to damage the coils, and it would be inconvenient to handle the crystals under such conditions. Any exterior means of heating, such as steam, boiling water, or hot oil, would involve too great a risk of breakage of containers and loss of valuable solutions. Large silica or porcelain basins have been used, and when the table below them is covered with sheet copper bent up at the sides and ends and brazed, so as to make a shallow vessel, such dishes can be used with reasonable safety. Their capacity, however, is too small for handling the quantities of material required at the plant of the National Radium Institute; therefore, it became necessary either to have larger vessels that would stand the action of boiling concentrated hydrochloric acid, or else to crystallize in neutral solutions. As crystallization in acid is much more rapid and efficient, the former method is much to be preferred.

A silica-lined, acid-proof ware made by Danto-Rogeat & Co., of Lyons, France, is exceedingly satisfactory. No other ware obtained either in this or any other country gave satisfactory results. In fact, most of the so-called “acid-proof ware” failed absolutely to withstand the acid. The institute was able to procure a number of small-size vessels from Danto-Rogeat & Co., and also, through the courtesy of the Welsbach Co., of Gloucester, N. J., to obtain one 250-liter steam evaporator made by the same concern. Owing to the European war, Danto-Rogeat & Co. were unable to supply any large-size vessels, as they had none in stock and their factory was closed. In September, 1915, they were about to start operations again, and an order for a number of steam evaporators has been placed for the purpose of equipping the plant as indicated in figures 2 and 3.

Distilled water is obtained for this part of the work by passing live steam through baffle 59 (fig. 2), and thence through a 1-inch block-tin coil contained in condenser 57, which consists of an ordinary wooden tank with the necessary inlet and outlet for a water flow. The block-tin pipe runs into earthenware distilled-water reservoir 58, which has a capacity of 107 gallons, and the water is piped to any part of the building desired.

In the sulphate building are the office and a small chemical laboratory, which is used mainly for qualitative determinations, titration of acids, etc., or any chemical work in which a quick result is desired. The main chemical control work and radium measurements, as well as the final refining of the radium, has been done in the laboratories of the Bureau of Mines. There is also a small storeroom, built of reinforced concrete with a steel door, in which the radium-barium sulphates and other valuable material can be stored, so as to eliminate fire risk. The building also contains a small double ball mill,


66, motor driven, for grinding the sulphates and mixing them with charcoal for later reduction in an oil furnace.

Power House

The 90-horsepower Kewanee locomotive-type boiler, used for the first plant alone, was purchased when the plant was on the experimental basis. When it was decided to build the second plant and continue work during the full period covered by the agreement with the Bureau of Mines, an additional 150-horsepower tubular Kewanee boiler was installed. The boiler house (Pl. II) is of brick, with concrete floor, a brick-and-iron partition separating the boilers from the compressor and pumps. The coal bin will readily hold about 2½ cars of coal.

The compressor (55, Pl. II) is 16 by 18 inches, capable of 140 revolutions per minute, and delivers about 650 cubic feet per minute at 40 pounds. The main vacuum pump 54, a 10 by 20 inch Rotrex pump, belt driven, is connected with the suction, filters used for leaching in both plants, and, with all pots connected, will give a vacuum of 20 to 22 inches of mercury. It is protected by means of a baffle, 47, filled with brick, over which a strong solution of caustic soda slowly circulates. The small pump 54a was the one originally used in the first plant, and was transferred to the powerhouse when the second plant was built. It is a No. 3 Nash vacuum pump; is chain driven, and will give a vacuum of 15 to 17 inches of mercury. It is connected with the suction filters under the settling tanks and also with the sulphate building, and is protected by a baffle (48, Pls. VI and VII).

Between the sulphate building and the extension of the new plant, is a small building with concrete floor and concrete walls for the storage of sodium nitrate. This building is connected by means of a belt elevator (Pl. XII, p. 60) with tho bin in which the sodium nitrate is bedded and sampled, and from thence the nitrate is taken to the stills.

Equipment for Sodium Nitrate Recovery

The sodium nitrate is pumped from the storage tanks (11 and 12, Pls. III, VI, and VII) in each plant by means of a Worthington pump, 4½ by 2¾ by 4 inches, into the two evaporators (43, Pl VI). These are made of 3/8-inch steel, one being welded and the other riveted, each being 6 feet by 5 feet by 24 inches, with a slope toward the center. Steam is used for evaporating, each tank containing a grid made of 1½-inch Byers pipe. Compressed air is used to hasten evaporation. The tanks are elevated so that the solutions can run by gravity into the crystallizing pans (44, Pl VI, and Pl. XI, A), which, are made of 1/8-inch. steel. Each pan is 10 feet long by 5 feet wide, 8½ inches deep at one end and 12 inches at the other, sloping toward the deep end so that it will drain readily. The pans are protected by a wooden roof covered with claterite.

Furnace Room

When the second plant was built, the original boiler house was converted into a furnace room. This is equipped with three Case oil furnaces (88, 84, 85, Pl. III), the blast being furnished by fans driven by 1-horsepower motors. The oil is supplied under pressure from a tank outside the building. The furnaces were specially built, two of them (24 and 25) holding cast-steel pots 18 inches high by 16½ inches wide and three-fourths of an inch thick (Pl. XI, B, p. 58). The other (28), a little larger in size, holds three No. 100 graphite crucibles, and is used in the reduction of the radium-barium sulphate. In the same room is a small ball mill (26, Pl. III), 2 feet 10 inches by 2 feet 6 inches, in which the sodium uranate is ground, either alone or with any other material that is desired in its treatment.

Detailed Description of Operations


Handling of Nitric Acid

The nitric acid is handled in carboys, and was originally weighed on scales at the foot of the elevator close to the sampling room. Since the completion of the acid plant (Pls. XII and IX, Pl) it has been weighed as drawn from the storage pots. With the weighed ore it is carried by means of the elevator to the upper landing of the first plant, from which there is a connection to the upper landing of the new plant, so that the same elevator serves both plants.

The spout of each leaching pot (fig. 1) is filled with freshly washed sand in order to protect the rubber stopper, and the acid is then dumped from the carboys into the leaching pots (Pl. IX, B), each of which contains sufficient distilled water so that when the pot has been filled it contains 121 pounds of 100 per cent nitric acid diluted to 38 per cent strength. Each day the foreman in charge of this work receives a slip indicating the weight of acid to be weighed for each pot and the volume of distilled water to be added.

Heating Nitric Acid

Live steam is run into the acid through a ½-inch glass tube connected by rubber tubing with a steam line. The steam in the line is passed through a baffle so as to eliminate impurities, and contains only traces of sulphates. As the distilled water is usually hot the


time of heating is not long, but steam is rim in until the temperature of the acid is at least 85° C.

The ore, ground to 20 mesh, is slowly added to each pot, the ore sacks resting on the wooden coverings (Pl. IX, B) on the pots, and the workmen using wooden paddles (Pl. IX, B) to stir the acid as the ore enters. Frequently calcium carbonate in the ore causes effervescence, and care has to be taken that the ore is not added so rapidly that frothing makes the acid overflow. Five workmen and a foreman can handle the 14 pots that are in the two plants and do all the necessary work connected with this part of the process. The heat of solution makes the temperature of the acid gradually rise during the addition of the ore, the temperature finally reaching 91° or 92° C. The addition of steam continues for 15 minutes after the last of the ore is in, the workmen stirring the different pots as rapidly as possible during this heating.

Nitric acid is not a very good solvent of vanadium, although it readily breaks up the vanadium minerals, even roscoelite. On the other hand, if too much vanadium is present, there is a tendency for vanadic acid to separate out in the acid as a reddish-brown precipitate. As this is rather gummy it retards filtration. If considerable vanadium separates in this manner, filtration may be so retarded that the acid cools down, causing a considerable loss of radium by reprecipitation. If there is little or no separation of vanadium in the acid, filtration takes place readily and quickly, and the extraction of the radium is thereby increased. Therefore, if the heating is continued with the object of decomposing the vanadium minerals other than carnotite, there is likely to be a loss of radium that will much more than compensate for the increased yield of vanadium. Consequently, it is much better not to continue the heating too long, but to make the extraction of the radium the main object. The uranium is practically all dissolved under the conditions mentioned.

Use of Hydrochloric Acid

The addition of a small amount of hydrochloric acid to the nitric acid increases the solvent action on the radium. Therefore, if the nitric acid does not contain hydrochloric acid, enough hydrochloric acid is added to make the proportion of concentrated hydrochloric acid to 100 per cent nitric acid, about 2.5 per cent. In other words, to each pot, 9 pounds of 28 per cent hydrochloric acid, the strength of the acid available in Denver, is added. This hydrochloric acid, of course, appears as sodium chloride mixed with the sodium nitrate, and the nitric acid made from the sodium nitrate contains some hydrochloric acid.

When heating has continued long enough, a wooden plug on the end of a long handle is inserted into the pot and pushed through the y ore until it almost completely stops the inside opening of the spout. By means of a lever (fig. 1) the rubber stopper is then removed from the lower part of the spout, and by raising the plug a little the acid is allowed to run onto the upper part of the suction filter (Pl. IX, C) below. In this manner, the flow of acid can be readily controlled, and at the same time a considerable part of the ore is retained in the leaching pot and receives the benefits of the acid wash. Thus, this method of handling the material has partly the effect of decantation.

Use of Acid Wash

An acid wash that has been prepared and heated in a small wooden tank, of which there is one in both plants (see 36, Pls. IV and VII), is run into each pot, and the heating, by means of live steam, continues. The acid flows by gravity through an acid-proof rubber hose, and each pot receives an acid wash of about 170 pounds of 10 per cent nitric acid. When the acid leach on the suction filters has gone through, the acid wash, with the remainder of the ore, is dumped onto the filters in the manner described above.

Washings with Distilled Water

When this material has completely filtered, the ore receives two washes of hot distilled water of 200 pounds each. Before the distilled water is added, however, the vacuum is broken on the filters and the liquid below, which consists of the acid and acid wash, runs from the bib cocks with which each filter is provided and flows through an earthenware pipe, which connects the different filter’s, to the radium precipitation tank (1, Pls. III, IV, VI, and VII). This tank contains 5 or 6 inches of water so that the hot acid is diluted enough not to damage the tank seriously. When the distilled water washes have filtered they are added in the same way to the acid liquor. Usually, the leaching process and filtration are complete between 2 and 3 o’clock in the afternoon. While the filtering is actually under way the workmen are weighing out the acid and ore for the next day’s run, and transporting it from the first floor to the third, so that leaching on the next day can begin promptly. As soon as filtration is complete the residue, consisting mainly of silica, is shoveled from the upper parts of the filters into a small trolley, which runs along the platform on which the filters are placed, and hauled to the dump.

Addition of Sodium Hydroxide

The acid solution in the radium precipitation tank (1, Pls. III, IV, VI, and VII), together with the wash water, is partly neutralized with sodium hydroxide, which is contained in small iron reservoir tanks (30, Pls. IV and VII). The tanks are filled as needed through pipes from the main sodium hydroxide reservoir (45, Pls. VI and VII), the liquid being propelled by means of compressed air. Sodium hydroxide is added at first rapidly and then slowly, the solution being stirred all the time. After a certain amount of the sodium hydroxide has been added a greenish precipitate appears in the tank. As more alkali is added this precipitate gets heavy and finally tends to turn slightly brown. At this stage the addition of the alkali should stop. As a rule, about 70 pounds of sodium hydroxide is required per ton of ore treated, but no set rule as to quantity can be given, as the iron and vanadium content of the ore varies. The more iron and vanadium the ore contains, the less sodium hydroxide can be added without carrying the precipitation too far.

Radium Precipitation

About 2 pounds of barium chloride, in solution, per ton of ore treated is then poured into the tank. As a rule the barium chloride is that recovered in fractionation and contains small quantities of radium. After the solution has been stirred for five minutes in order that the barium chloride may be well mixed in, sulphuric acid is added, about 15 pounds of 100 per cent acid per ton of ore treated. The sulphuric acid not only precipitates barium sulphate, which carries down the radium, but also dissolves the small precipitate of iron and vanadium in the solution. Consequently the final precipitate obtained is a fairly clean one of radium-barium sulphate, containing only small traces of iron, vanadium, etc.

Should the addition of the sodium hydroxide be carried too far, the precipitate obtained is what is called a “retreat”; that is to say, it carries more iron and vanadium than is advisable. Under such conditions the wet sulphate is placed in an earthenware pot (50, Pl. VII) and concentrated sulphuric acid is added, the whole being thoroughly stirred. Water is then run in and the iron and vanadium readily dissolve in the hot liquor obtained, leaving the radium-barium sulphate as a clean white residue. The liquid and residue are run through a rubber hose to one of the filters (32, Pl. VII). Each filter is used only one day out of four for filtering the regular precipitate from the settling tanks (Pl. X and 31, Pl. VII). The amount of sulphuric acid required for leaching a retreat varies according to the amount of iron and vanadium in the precipitate, but usually 25 to 35 pounds of 66° B. acid suffice. The acid filtrate obtained is used as part of the sulphuric acid required for precipitating the next batch of radium-barium sulphate in the precipitation tank (1, Pl. VII).

After the precipitation of the radium-barium sulphate, stirring is continued for one hour, when the whole solution is pumped into one of the settling tanks (31, Pl. VII) through the centrifugal pump (15,
Pls. III, VI, and VII). In the first plant the settling process occupied three days, as there were only three settling tanks; in the second plant the liquids are allowed to settle four days, there being four tanks. On the second day after the liquid has been run into the tank a solution containing 1 pound of barium chloride is poured in and thoroughly stirred into the upper part of the acid liquor by means of a paddle.
The object is to carry down any small traces of radium that may have remained suspended or dissolved in the supernatant liquid.

Precipitation Processes

Iron precipitation by use of sodium carbonate

When the settling is complete the clear liquid above the precipitate is siphoned off through a 1¼,-inch acid rubber hose attached to a floating siphon. The acid liquor is run into tank 3 (Pls. III, VI, and VII), which contains a hot solution of sodium carbonate more than sufficient to neutralize the acid. The weight of sodium carbonate used depends to some extent on the character of the ore treated. The smaller the amount of uranium and vanadium in the ore, the smaller the excess of carbonate required. For ore running from 2.5 to 3 per cent U3O8, 250 pounds of excess sodium carbonate is required per ton of ore treated. This figure does not take into consideration the addition of the sodium hydroxide in the radium precipitation tank, but is calculated on the original acidity of the acid used. Consequently if acid of the strength and in the quantity described under the section on leaching is used the actual weight of sodium carbonate used is about 650 pounds of soda ash per ton of ore. In reality the excess of sodium carbonate is larger than that indicated, owing to the fact that a part of the acid is neutralized by the sodium hydroxide added in tank 1. A 2 per cent U3O8 ore would not need more than 200 pounds excess sodium carbonate per ton of ore unless it contained more than the usual amount of vanadium.

The acid liquor is siphoned into the hot sodium carbonate, the whole process taking about three hours. As long as the alkali is in excess, there is little chance of the solution boiling over; should the neutralization point be nearly reached, there is serious danger of Such boiling. The liquid is heated during the addition of the acid, and maintaining the temperature near the boiling point during the entire period is important; otherwise the precipitate, which should be red, becomes brown or even bluish-brown, and carries considerable vana-



dium and too much uranium. Not only is the amount of sodium carbonate in excess important but also the time of boiling after the acid has been completely added. In actual practice, the boiling is continued for three hours after the last of the acid has been run into the tank. There are, therefore, two important factors that control the amount of uranium and vanadium left in the precipitate, namely, the excess of sodium carbonate and the period during which boiling is continued, the latter being just as important as the former.

Filtering Through Presses

The solution is then filtered through filter presses—in the first plant through press 33a (Pl. IV); in the second plant through press 16 (Pls. VI and VII), the filtrates running into tanks 6 (Pls. III, VI, and VII), in which the uranium is afterwards precipitated. The precipitates are washed in the filter press for about 20 minutes, the washing being added to the original solution. It is, of course, advisable to wash the precipitates until they are as free as possible from adhering liquor so as to decrease the loss on uranium and vanadium, but the amount of washing is controlled by the capacity of the tanks into which the filtrate is run.

The uranium and vanadium content of the iron precipitate obtained has varied considerably, the average uranium content running about 0.7 per cent U3O8, and the average vanadium content about 2 per cent V2O5. Some of the precipitates have gone as low as 0.5 per cent U3O8 and 1 per cent V2O5, whilst once or twice, when there was a temporary lack of steam and the boiling was not continued sufficiently long, the uranium content of the iron precipitate was more than 2 per cent U3O8.

Recovery of Barium Sulphate Precipitate

After the acid liquor has been added to the sodium carbonate tank, as described above, and before the filter-pressing of the iron precipi-tate, the radium-barium sulphate precipitate in the conical settling tanks is run through earthenware stopcocks at the bottom of each settling tank onto the suction filters. The solution is filtered under suction, and, by means of a rubber hose attached to the spout of the filter, the filtrate is added to. the sodium carbonate tank. The sulphate is washed with distilled water, and with a weak solution of sodium hydroxide, in order to neutralize any adhering traces of acid. It is then carefully scraped from the filter paper placed over the asbestos filter cloth and is put into an iron pan 19 by 20 by 3 inches. The sulphate is dried in a hot-air oven heated by steam coils, after which it is transferred to the sulphate building for further treatment. If the solution at any time runs through cloudy, owing to the presence of barium sulphate in the filtrate, it is run back into tank 1 (Pls. III, IV, VI, and VII) by means of a long rubber hose, and the sulphate that has gone through the filter is thus recovered with the next radium-barium sulphate precipitate. It is found preferable to do this rather than to try to refilter the solution at once. Such mechanical losses in plant operation should be carefully watched, as they are more likely to occur than chemical losses.

Re-Solution of Iron Precipitate

When the first plant was started it was thought that not only would the recovery of the uranium and vanadium content of the iron precipitate be justified, but also that the radium not recovered in the first radium-barium sulphate precipitate might be sufficiently large to justify its recovery also. Therefore, the original plan of procedure involved the re-solution of the iron precipitate in hydrochloric acid and the precipitation of a second radium-barium sulphate precipitate, followed by a second precipitation with sodium carbonate.

Accordingly, into tank 2 in the first plant was poured a sufficient amount of commercial hydrochloric acid to just dissolve the iron precipitate. This was diluted with an equal volume of water. The amount of acid required varied with the ore treated. Some ores carry more iron than others, and, therefore, the exact quantity of hydrochloric acid required in all cases can not be stated. In addition, if the iron precipitate carried a little more vanadium than usual, this affected the quantity of acid required.

Generally speaking, however, 950 pounds of 28 per cent hydrochloric acid was required for each ton of ore treated. The acid was used cold, the iron precipitate being added slowly, one shovelful at a time, the liquor being thoroughly agitated by a paddle during the addition. If the iron precipitate is added too rapidly, the solution heats, and toward the end of the reaction there is a tendency for the vanadium to separate. Under these conditions it is difficult to get the whole mass in solution, so that when the radium-barium sulphate is precipitated, an unsatisfactory, dirty sulphate is obtained.

When all the material is in solution the tank is almost filled with water and 3 pounds of barium chloride added. This is followed by the addition of 30 to 40 pounds of 66° E. sulphuric acid. The dilution and the amount of sulphuric acid added is affected, to some extent, by the amount of calcium in the ore, as calcium sulphate is likely to be precipitated if the solution is not sufficiently dilute.

The radium-barium sulphate thus obtained was elevated to one of the settling tanks and allowed to settle until the next day. The clear liquid obtained was siphoned off by means of a floating siphon, in the ordinary way, and run into an excess of hot sodium carbonate, the iron-calcium precipitate obtained being filter-pressed, and the uranium and vanadium in the filtrate recovered.

It was found that the cost of recovering this uranium and vanadium was greater than the value of the products obtained, and after a few trials, the second precipitation of uranium and vanadium was abandoned, and the tanks that were designed for use in this part of the process, namely, tanks 2, 4, 5, and 7 (Pls. III, IV, B, and V) were afterwards used for other purposes. However, the second precipitation of the radium was continued for a longer period of time, “second” sulphates being obtained on the first eight cars of ore treated. Later, it was found that the recovery on one car (P-7) was not satisfactory, and that some of the radium was going through in the filtrate from the first radium-barium sulphate precipitate. The re-solution of the iron precipitate and the formation of second sulphates was again started at this time and continued with the ore from three cars (P-7, P-8, and P-9). As the cause of the losses was then ascertained to be of a mechanical nature, and eliminated, the formation of the second sulphates was again discontinued, as the amount of radium obtained in the sulphates did not justify the expense.

Precipitation of Uranium

The filtrate from the iron precipitate carries the uranium in solution as a double sodium uranyl carbonate and the vanadium as sodium vanadate. The solution is partly neutralized, acid being added until a yellow precipitate begins to form. This precipitate is supposed to be uranyl carbonate, but in reality it contains more sodium uranate than uranyl carbonate, as on ignition it gives a very small amount of oxide. If too much acid is added at this stage the amount of vanadium that appears with the sodium uranate is considerably increased. When the precipitate begins to foam the addition of acid is stopped and sodium hydroxide is added to the hot liquor until the uranium is completely precipitated as sodium uranate. The completeness of the precipitation is easily tested by filtering off some of the precipitate on a small funnel and adding more sodium hydroxide to the filtrate in the test tube or beaker.

During the operation of the first plant, before the second had been built and before a nitric-acid plant had been erected, sulphuric acid was used to partly neutralize the sodium carbonate. The sulphuric acid was run into the iron tank (13, fig. 4) and elevated by compressed air into the lead-lined tank (41, fig. 4), where it was diluted and run through lead pipes to the uranium-precipitating tank (6, Pl. III) and the vanadium tanks (7, Pl III, and 89, fig. 4 and Pl V). Therefore the filtrate from the vanadium precipitate afterwards obtained contained not only sodium nitrate but also large quantities of sodium sulphate, and it was necessary to separate the sodium sulphate by means of fractional crystallization, as described subsequently. After the nitric-acid plant had been erected it was found much more satisfactory and, in the end, cheaper to neutralize at this point with nitric acid instead of with sulphuric acid, so that the grade of sodium nitrate obtained could be greatly increased and fractional crystallization could be eliminated.

After the addition of the sodium hydroxide and the precipitation of the uranium the solution is boiled for one hour in order to promote complete precipitation and is then filtered through press 17 (Pl. VI). This press takes care of the uranium in both plants, although origin-

section of first or old plant

ally the uranium precipitate in tank 6 (Pl. III) in the first plant was filtered through press 33c (Pl. IV, A). The sodium uranate cake is washed for about 15 minutes and then dried in pans in hot-air ovens (22, Pls. V and VII). This method of drying has been found to be the best under the circumstances, not only for the radium-barium sulphate precipitates, but also for the uranium and the vanadium.

Trial was made of another method, embracing the use of large iron pots heated by direct heat from a small fire placed beneath, but the results were not satisfactory; both the uranium and the vanadium precipitate caked readily and dried slowly, notwithstanding considerable hand stirring. The method, therefore, was abandoned.