|FaradNet The Capacitor Resource|
|Home | Community | Companies | Catalogs | Knowledge | Search | Store|
The chemical cleaning of plain surface anodic material requires a certain technique but this does not suffice for the proper cleaning of etched surface material. For that reason, cleaning methods will be divided into those applying to plain and etched surfaces.
As the majority of wet electrolytic capacitors employ anode assemblies fabricated from either hard or soft aluminum sheet or foil varying in thickness from 0.003" to 0.010", cleaning methods will be confined to anode material of that nature.
Aluminum foil comes from the mill in the form of rolls and the surface of the material is perfectly smooth with a bright mirror-like finish, but the surface is generally covered with a thin film of oil or other lubricant which has been used in the rolling mill operations. This is particularly true of hard foil which obviously does not pass through any annealing operation. Soft foil which has been annealed after the last rolling mill operation, as a rule, has lost almost all of the surface lubricant by vaporization due to the heat of annealing.
Some fabricators of aluminum foil incorporate washing operations in lubricant solvents to remove such surface materials but all aluminum foil, intended for use as anode members, in wet electrolytic capacitors, must be thoroughly cleaned before the dielectric film is formed.
Considerable improvement has been noted in recent years, however, in the surface condition of aluminum foil, fabricated especially for use in electrolytic capacitors. This improvement has, in the main, been obtained by the elimination of the use of mineral oils and the substitution of vegetable oils such as palm or cocoanut oils as the lubricating mediums in rolling mill operations. As the vegetable oils will saponify readily with various alkali cleaning agents much better surface cleaning is therefore obtained.
Book ToC | Book Index | Page Index
For simple cleaning of the aluminum surface with only mild etching trisodium phosphate (Na3PO4) is a comparatively effective reagent. Here also the treatment must be followed by an immersion in nitric acid.
The etching action of alkalies, particularly the phosphates and carbonates, can be minimized without interfering with the cleaning action by adding an inhibiting agent such as sodium silicate in an amount equal to approximately 5 per cent of the weight of alkali used.
In some cases the aluminum surface may be coated with a layer of mineral oil that cannot be readily removed with the caustic solutions. In such cases it may become necessary to first wash the material in such solvents as benzol, alcohol or trichlorethylene as a preliminary treatment.
One excellent cleaning method involves the immersion of the aluminum in a solution consisting of concentrated sulphuric acid and potassium dichromate. Such a combination affords an extremely strong oxidizing reagent which has no appreciable effect on the aluminum surface as long as the water content of the solution is negligible. This solution effectively removes all carbon compounds and at the same time seems to coat the surface of the metal with a very thin protective film of desirable characteristics.
Of all the cleaning methods known, the potassium dichromate-sulphuric acid one produces the most desirable results.
Apparently, there are three basic types of cleaning procedures which produce comparative results and illustrative of these are the following block diagrammatic presentations.
In the cases where the use of sodium hydroxide is mentioned, it is interesting to note that considerable latitude is encountered in the selection of desirable concentrations, temperatures and times of immersion. In practice, concentrations vary from 1% for 5 minutes at 95°C; 3%, for 3 minutes at 80°C; to 10%, for 2 minutes at 40°C.
The potassium dichromate-sulphuric acid generally consists of 30 grams of finely ground potassium dichromate, dissolved in each liter of concentrated 660 Baume sulphuric acid.
After the final cleaning or washing operation by any of the mentioned methods, anode assemblies or foils are ready for the operation of being anodically filmed or formed with the dielectric coating. Such filming or forming operations take place immediately following the final cleaning or washing operations.
From a study of the foregoing procedures it becomes immediately apparent that anode assemblies, prior to anodic film forming operations, must be very thoroughly cleaned. It should also become equally apparent that once the cleaning operation has been completed, anode assemblies cannot be handled or even touched with the bare hand or other contaminating agents.
Anode assemblies using plain surface aluminum foil are always mechanically fabricated into finished form with riser rods or stems riveted or eyeleted in place. The entire assembly is then cleaned as an integral unit. Riser rods and rivets are as a rule given no preliminary cleaning other than a wash in solvents such as benzol, alcohol or trichlorethytene.
Anode assemblies, on the other hand, which incorporate the use of etched aluminum foil, must be etched prior to the attachment of the riser rods because riser rods and fastening rivets or eyelets are not etched for apparent reasons. This necessitates a different washing and cleaning technique for etched foil anode structures prior to the forming of the anodic dielectric film.
Book ToC | Book Index | Page Index
Chlorides are removed by washing in tap and distilled water washes and heavy metals exposed by the etching process or plated on the aluminum surface are removed by immersion in nitric acid. The use of the nitric acid obviously also necessitates the use of repeated washes to remove all traces of nitrates.
It is found to be very desirable to remove all traces of these contaminating substances before the etched foil is attached to the riser rods. This is nearly always the procedure.
In riveting, eyeleting or otherwise attaching anode foils to riser rods, the etched and washed foil must be handled so it becomes necessary to again wash the completed anode assembly thoroughly before formation.
A number of typically basic cleaning procedures for etched anode foil assemblies are outlined in the following block diagrams:
In the cleaning of all types of anode assemblies it has been observed, from experience, that washing of aluminum surfaces in hot water is to be avoided as much as possible for the following reasons. Aluminum when immersed in water that is neutral to pH7 or slightly alkaline hydrolyzes to form and take on a coating of aluminum hydroxide and hydrogen is evolved. Such coatings of aluminum hydroxide are very nearly impossible to remove and subsequently retard anodic film formations as well as affect very materially the contact resistance between the oxide film and electrolyte, in completed capacitors. This high resistance layer of aluminum hydroxide manifests itself in the form of increased equivalent series resistance of a capacitor and a heavy layer of aluminum hydroxide may in some cases entirely prevent a satisfactory anodic film formation.
Acidifying wash waters slightly with boric acid tends towards reduction of the amount of aluminum hydroxide as does the use of strong nitric acid and p0tassium dichromate-sulphuric acid solutions. It has been further observed that exposure of freshly cleaned aluminum surfaces to warm or heated moist air also tends towards the formation of the unwanted aluminum hydroxide layers.
Chemically etched surfaces are obviously more susceptible to this condition than plain surfaces and great difficulties are frequently encountered in the processing of etched foil capacitors.
A number of methods for drying etched foil quickly and without the application of heat, in order that aluminum hydroxide formation may be reduced, include centrifugal drying and drying with higher alcohols of the hydroscopic class.
Book ToC | Book Index | Page Index
Experience has shown that there are two distinct types of films: a thin film of aluminum oxide (A1203) and a comparatively thick film of hydrated aluminum oxide or aluminum hydroxide which also may contain other elements such as aluminum sulphate, oxalate, chromate or other acid reaction products.
The type of electrolyte used will determine, to a great extent, the type of film produced but there are other factors such as temperature and magnitude of applied electrical potential.
Experimental determinations have demonstrated the fact that anodic films can be formed in the followmg list of electrolytes:
This is, by no means, a complete list of the possible electrolytes which have shown results but it will readily serve to show the trend.
Theoretically, the ideal electrolyte would be distilled water in order that only oxygen would be evolved at the anode surface. Obviously this would be impossible and the water must be rendered more conductive by the addition of other substances, in order that current of sufficient magnitude for the desired electrolysis can be conducted.
If aluminum is immersed in an aqueous solution of sulphuric, phosphoric or oxalic acid and made the anode, it will be found, upon passage of current, that the aluminum surface will acquire a tough film of a rather gelatinous nature. It will also be found that the same magnitude of current will pass with time. In other words, the film does not show a current limiting action as it is built up. This demonstrates that the film will continue to be increased in thickness as long as voltage is applied. It has been noted that the thickness of such a film is determined by the time of current application and the acid concentration. The acid concentration obviously determines the current density or magnitude at any given polarization potential. The operative polarization potential in this type of film formation is definitely limited to a narrow range and is different for each of the three acids mentioned. Illustrative of these anodic film formation methods are the following actual procedures:
This type of film does not possess the asymmetric characteristics of the aluminum oxide film and is therefore known as an inactive film to distinguish it from the extremely thin dielectric film of aluminum oxide known as the active type of film. The inactive type of film is apparently a form of hydrated aluminum oxide which contains some aluminum salts of the acid used in the electrolyte. Although not a satisfactory dielectric, the inactive film may be rendered conductive in the presence of alkali ions, which fact may permit the subsequent formation of an active oxide film on the aluminum surface, beneath the inactive film. The presence of an inactive film, however, reduces the effective capacity and increases the equivalent series resistance of a completed capacitor of either the wet or dry electrolytic type. On the other hand, the inactive film is sometimes used as a mechanical protective coating for the active oxide film. This will be referred to again in later chapters.
The true active dielectric film of aluminum oxide is best formed in an aqueous solution of ammonium or sodium borate and the concentrations of the borate salt are desirably such that the electrolyte is sufficiently conductive but definitely on the acid side of neutrality. To accomplish this, boric acid is added to the electrolyte. As the phenomenon of sparking or voltage breakdown of the active anodic film is a function of the log of the ion concentration and temperature of the electrolyte, the net result is that for satisfactory anodic film formations the electrolyte consists, in practice, of an aqueous solution of boric acid with a comparatively small content of either ammonium or sodium borate.
In practice the boric acid content ranges from ten to sixteen per cent, by weight, of the total electrolyte solution.
Experience has shown that sodium borate has a number of practical advantages not possessed by ammonium borate and for that reason it is in general use.
The amount of sodium borate that is permissible in an electrolyte is governed by the voltage of film formation. The higher the voltage of film formation, the lower the permissible quantity of sodium borate. In fact, for formations above certain voltage ranges, the sodium borate content is reduced to zero.
Illustrative of the variation of permissible sodium borate content is the following tabulated data:
|100||15||--||650 to 700|
|100||15||0.12||550 to 600|
|100||15||0.25||500 to 525|
For lower formation voltages it becomes obvious that the sodium borate content may be increased in proportion to the reduction in forming potential but experience has shown that a thinner and more active film is produced with the lowest concentrations of sodium borate. For that reason the higher voltage electrolytes are as a general rule employed for forming oxidc films at practically all voltages.
It appears that the use of sodium borate contents above certain concentrations tends to result in the active film being covered with a layer of aluminum hydrate. Whether this is the case or it is that the active film is actually hydrated to some extent is not definitely known. Either one or the other of the two possible conditions results in an increase in contact resistance between the active film and the electrolyte, in finished capacitor structures.
In practice anode assemblies are either formed by one or the other of two processes called still and continuous formations.
In the still formation process, anode assemblies are immersed in electrolyte contained in copper lined wooden tanks. Voltage is applied between the copper lining and the anode. A low voltage is first applied and gradually and continuously increased in value while the value of current is maintained at some predetermined, constant value until the desired value of voltage is reached. When the desired value of voltage has been reached it is held constant until the current decreases to a predetermined minimum value. The formapion at this point is complete and the anode assemblies are removed, then assembled into capacitors.
In the continuous formation process, anode assemblies are carried by a mechanical conveying system in such a manner as to pass, at a uniform rate, through the electrolyte. In this process full desired formation voltage is applied at all times and the current allowed to decrease as the anode is moved through the electrolyte. The speed of travel is so adjusted that by the time the anode has traveled a predetermined distance, the current will have decreased to the value required for a complete film formation.
Although many advantages are claimed for one process versus the other, both processes produce equally satisfactory results although the still process of formation is in more general use than the continuous.
As a general rule, electrolytes are maintained at or near their boiling points during the formation of anode assemblies and the chemicals used to make up the electrolytes are of the highest purity obtainable. In this respect it is highly essential that electrolytes be kept free of impurities, such as chlorides, nitrates, sulphates, and iron, to the extent of less than one part in a million.
The following graphical illustration shows the relationship between current, voltage and time during a typical still formation cycle of wet electrolytic anode assemblies.
From expenence it has been found that satisfactory film formations require initial current densities of from 15 to 20 milliamperes per square inch of anode surface, for plain surfaces, and from 30 to 50 milliamperes per square inch of anode surface for etched surfaces. Anode assemblies which have acquired the desired oxide film show, as a general rule, a terminal or finished, steady state current density varying from 50 microamperes per square inch of plain anode surface to 100 microamperes per square inch of etched surface.
A number of variable factors are encountered in the formation of anodic films. It is found that the rapidity of formation varies to a considerable degree with the purity of the aluminum used in the anode assemblies. The higher the purity the more rapid is the film formation. This naturally follows because any presence of heavy metals such as iron or copper would prevent film formation at the points where they occurred as these heavy metals have no film forming characteristics. Even when present in the most minute degree the least effect that can be anticipated is an increase in the steady state current density or what is termed leakage current.
The presence of chloride contamination in the electrolyte or on the surface of the anode, even though the quantity be only a few parts in a million, will effectively decrease the rapidity of film formation or prevent formation at all due to the highly corrosive action of the chlorine ion on aluminum. So important is the question of chloride contamination that it is customary to check formation electrolytes almost hourly for the presence of chlorides. Other types of contamination are almost equally detrimental but the chlorine ion is the worst of all.
Improperly cleaned and aluminum hydrate coated anode surfaces very materially "slow up" the process of film formation.
One particular phenomenon has been noted in relation to the effect, on film formation, of the presence of aluminum hydrate or aluminum hydroxide coatings on the anode surface. Such coatings or layers are comparatively porous and therefore tend to entrap minute bubbles of oxygen during the formation process. These bubbles of oxygen tend to block current flow and the effect is a decrease in forming current seeming to indicate proper progress in the formation of the anodic oxide film. In many cases, forming current decreases at such a rate that an anode is considered to be completely formed when such is not the case but the current has been completely blocked by the occluded bubbles of oxygen. Such a condition can be checked by removing the applied potential for sufficient time to allow the oxygen to escape and be replaced by electrolyte and upon again applying voltage the current density increases to values normal to the actual condition of the oxide film.
At times, the elimination of aluminum hydrate films is so difficult that special formation procedures are necessary to offset the lowering of formation efficiency. Such a procedure sometimes consists of applying formation voltages in steps of, for example, 50 volt increases with one minute idling or zero potential periods in between until the final desired value of voltage is reached. Even after full voltage values are reached the circuit is opened at fixed intervals, during the forming cycle, for one to five minute idling periods. The sudden application of potential after an idle period of this nature is sometimes called "surging." Chemically etched anode assemblies are extremely affected by this detrimental coating of aluminum hydrate because of the nature of the etched surface initially.
After an anode is completely formed it is removed from the formation electrolyte, rinsed in cold distilled water or cold distilled water with a boric acid content of not more than three per cent. It is now ready to be assembled into the container and the capacitor completed but as a test on the effectiveness of the anodic film it is customary to wait from twelve to twenty-four hours and then test the anode for leakage current. Such a test is made by immersing the anode in cold electrolyte of the same chemical composition as the electrolyte intended for use in the completed capacitor, applying voltage and reading the current which the oxide film passes at the end of one minute.
(Courtesy Cornell-Dubilier Electric Corp.)
Book ToC | Book Index | Page Index
The following graph indicates surface areas of the anode surface but by this it is meant one side of the anode foil or plate and not the actual total surface area, which would obviously include both sides. All previous and future references of such a nature will be made on the same basis.
Book ToC | Book Index | Page Index
Fill electrolytes are aqueous solutions of boric acid and sodium or ammonium borate. The use of ammonrum borate is general today where aluminum containers are almost exclusively used for wet electrolytic capacitors. Some nine or ten years ago copper containers were in general use and at that time sodium borate was used, the ammonium borate being obviously not suited for use in such containers.
Wet electrolytic capacitors are primarily designed to operate at normal room temperatures; thus a deter-mining factor for the quantity of boric acid, which can be used in the electrolyte, is automatically established. The maximum quantity of boric acid which can be used is limited to that amount which will remain in solution at room temperature, in other words 20° centigrade, which temperature is nominal low value. Experimental determination shows this quantity to be a maximum of 5 grams per 100 c.c. of water.
The quantity of ammonium borate in the fill electrolyte or rather the quantity of ammonia, is governed by two factors; namely, the minimum rated sparking or scintillating voltage in one case and the necessity of keeping the electrolyte definitely on the acid side of neutrality in the other case. With these limitations, fill electrolytes are confined to a fixed range of chemical composition.
While theoretically it would appear that the phenomenon of dielectric film breakdown would occur over a wide range of applied potentials, such is not actually the case and audible scintillation or visible sparking rarely occurs below a potential of 350 volts even though the ion concentration of the electrolyte be relatively high.
As ion concentration is the determining factor of electrolyte conductivity it is customary to determine ion concentrations by simple resistivity measurements.
The following graph shows the relation between electrolyte resistivity and variable ammonia content. As temperature is also a determining factor of resistivity, the same graph also shows the effect of temperature variation. The graph is based on a fixed boric acid content of 5 grams and a water content of 100 c. c. for reasons already mentioned.
A study of this data reveals two important facts. One is the variation of specific resistivity with ammonia content and the other is the fact that the lower the ammonia content, the greater the change in specific resistivity with a given change in temperature. It can furthermore be seen that the higher the operating voltage of a wet electrolytic capacitor the higher its equivalent series resistance and the greater its change in equivalent series resistance with change in temperature.
As the scintillating voltage is a log function of the specific resistivity of the electrolyte, temperature and dielectric film thickness it must hold true that scintillating or breakdown voltage varies with the ammonia content. To illustrate the relation between these values reference is made to the following graphical illustration.
The values of scintillating voltage shown are based on measurements made at 20 C. with a dielectric film formed initially to 525 volts.
The wet electrolytic capacitor utilizing an aqueous electrolyte is obviously not adapted for operation in ambient temperatures near the freezing point of water because near that temperature the electrolyte will also freeze. To overcome this limitation electrolytes, incorporating the addition of a polyhydric alcohol such as glycerol or ethylene glycol, have frequently been employed. With the use of such an electrolyte, the operating temperature range has been extended to values as low as minus 30 degrees centigrade with satisfactory results. Such an electrolyte, with a scintillating voltage (at 20°C) of 500 volts, is shown below:
One serious disadvantage to the employment of such ah electrolyte has been noted and that is, should the potential applied be of such magnitude to cause scintillation or sparking at the anode surface, the heat of the spark will cause decomposition of the polyhydric alcohol and small deposits of carbon will appear on the anode surface. Such deposits of carbon will increase the leakage current of the capacitor and may, if the deposits are large, render the capacitor unsatisfactory for further use.
It has been found that glycerol is much more susceptible to such decomposition than is ethylene glycol.
(Courtesy Cornell-Dublilier Electric Corp.)
Book ToC | Book Index | Page Index