CA1079350A - Fuel cell using iron chloride electrolytes - Google Patents
Fuel cell using iron chloride electrolytesInfo
- Publication number
- CA1079350A CA1079350A CA291,579A CA291579A CA1079350A CA 1079350 A CA1079350 A CA 1079350A CA 291579 A CA291579 A CA 291579A CA 1079350 A CA1079350 A CA 1079350A
- Authority
- CA
- Canada
- Prior art keywords
- anolyte
- catholyte
- energy conversion
- conversion apparatus
- accordance
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Expired
Links
Classifications
-
- Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
- Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
- Y02E—REDUCTION OF GREENHOUSE GAS [GHG] EMISSIONS, RELATED TO ENERGY GENERATION, TRANSMISSION OR DISTRIBUTION
- Y02E60/00—Enabling technologies; Technologies with a potential or indirect contribution to GHG emissions mitigation
- Y02E60/10—Energy storage using batteries
Landscapes
- Hybrid Cells (AREA)
Abstract
ABSTRACT OF THE DISCLOSURE
An instantly refuelable energy conversion system com-prises a multicell array wherein each cell has a cathode, a solid anode fuel selected from the class consisting of iron, zinc and nickel and a ferric (and/or +3 valence nickel or zinc) chloride liquid oxidizer. In power delivery (discharging) e.g., using iron alone, the reactions are:
At the anode, Fe?Fe+++ 2e yielding 0.44 volts, At the cathode, Fe++++e?Fe++ yielding 0.75 volts, the overall reaction being:
Fe + 2FeCl3?3FeCl2, i.e., 1.2 volts overall.
These reactions are reversible for electrical charging purposes The oxidizer is recirculated continuously and chemical reaction between the oxidizer and solid fuel in the fully charged or partially charged system is limited by a diffusion barrier be-tween the oxidizer and fuel so that reactions therebetween are predominantly electrochemical. The system is characterized by continuously countering the tendency of rising pH of anolyte, and/or of contaminant accumulation therein, to maintain stable operation over long periods, e.g., by using a main energy con-version system anolyte as a catholyte in an auxiliary, over-driven energy conversion system which at the same time the catholyte of the main system serves as anolyte of the auxiliary system with reliance on the richness of the liquid in hydrogen ion and Fe+++ ion to chemically attack any elemental iron formed in the auxiliary system thereby further preserving stability.
The iron may be substituted in whole or in part by zinc and/or nickel.
An instantly refuelable energy conversion system com-prises a multicell array wherein each cell has a cathode, a solid anode fuel selected from the class consisting of iron, zinc and nickel and a ferric (and/or +3 valence nickel or zinc) chloride liquid oxidizer. In power delivery (discharging) e.g., using iron alone, the reactions are:
At the anode, Fe?Fe+++ 2e yielding 0.44 volts, At the cathode, Fe++++e?Fe++ yielding 0.75 volts, the overall reaction being:
Fe + 2FeCl3?3FeCl2, i.e., 1.2 volts overall.
These reactions are reversible for electrical charging purposes The oxidizer is recirculated continuously and chemical reaction between the oxidizer and solid fuel in the fully charged or partially charged system is limited by a diffusion barrier be-tween the oxidizer and fuel so that reactions therebetween are predominantly electrochemical. The system is characterized by continuously countering the tendency of rising pH of anolyte, and/or of contaminant accumulation therein, to maintain stable operation over long periods, e.g., by using a main energy con-version system anolyte as a catholyte in an auxiliary, over-driven energy conversion system which at the same time the catholyte of the main system serves as anolyte of the auxiliary system with reliance on the richness of the liquid in hydrogen ion and Fe+++ ion to chemically attack any elemental iron formed in the auxiliary system thereby further preserving stability.
The iron may be substituted in whole or in part by zinc and/or nickel.
Description
3~V
BACKGROUND o~i T~IE INVENTION
The present inventlon re:La-tes to energy conversion and more par-ticularl~ to a rechargeable and/or refuela~le high specific power delivery and high specific energy s-torage sys-tem A large number of por-table power delivery and energy storage electrical devices inc]uding automobiles, golf car-ts, forklift trucks, garden appliances, cons-truction equipmen-t, portable motors and lighting uni-ts, advan-tage of Fe/~e 3 here require long life and reliable means for high specific power delivery and high specific energy s-torage capability ("specific"
meaning with respec-t -to weigh-t). ~ow cost and long life energy storage systems are needed in -the above applications and also in connecti.on with data processors and instrumen-ts and load levelling The state of the art, including the most promising of current R&D efforts, is summarized in Iammartino, "New Batteries Are Coming", Chemical Engineering magazine, pp ~8-50, January 20, 1975, The research and development effor-ts include use of the electrochemical pairs zinc-nickel oxide, metal-air (oxygen), sodium-sulfur, li-thium-sulfide, zinc-chlorine and my own prior efforts with iron-ferric chloride.
It is an important object of the present invention to provide such a system free of -the disadvantages of prior systems I-t is a further object of the present invention to provide low cos-t of initial capital equipmen-t consistent with ;~; the preceding object, ; It is a further object of the invention to provide low operating costs consistent with one or more of the pre- ~ -ceding objects.
It is a further object o~ the invention to provide indefinitely long life of capital equipment consistent with ; one or more of the preceding objects.
; ~ ':' :, -- 2 -- .
'793~i~
I-t is a fur-ther objec-t of -the invention to provide subs-tan-tially hazard-free opera-tion and handling consistent with one or more of the preceding objec-ts.
It is a further object of -the inven-tion to provide capabili-ty of being "refueled" immediately consistent with one or more of the preceding objects.
It is a further object of the inven-tion to provide capability of being recharged electrically consis-tent wi-th one or more of the preceding objects.
It is a further object of the invention -to avoid -the necessity for a complex chemical supply network or logis-tics consistent with one or more of the preceding objects.
It is a further object of the inven-tion -to provide odorless operation consistent with one or more of the preceding objects.
It is a further object of the invention to provide low freezing temperatures consistent with one or more of the preceding objects, It is a further objec-t of the invention to provide for complete discharge wi-th no damage consistent with one or more of the preceding objects.
It is a further object of the invention -to provide simple means of indicating remaining energy consistent with one or more of the preceding objects.
It is a further object of the invention to provide -indefinitely long charge retention consis-ten-t with one or more of the preceding objects.
It is a~further object of -the inven-tion to provide that when the system is turned off, there is no power available at the terminals, thus making the system electrically safe to handle when "deactivated" conslstent wi-th one or more of the preceding objects, - ;~
It ls a further object of -the inven-tion -to provide ex-tremely plen-tiful and non-pollu-ting ma-terials ernployed con-sistent with one or more of -the preceding objects.
It is a further objec~ of the inven-tion -to provide an energy conversion device ~or which simple manu~acturing methods can be employed consisten-t with one or more of -the pre-ceding objec-ts.
SUMMARY OF THE INVENTION
According to the inven-tion, an energy conversion system is made in a fashion to be rechargeable in si-tu as a secondary battery or, alternatively, ins-tan-tly refuelable in the fashion of a solid li.quid fuel cell and comprises a ca-thode and, when charged, a solid anode fuel comprising iron, zinc and/or nickel and a liquid oxidizer comprising ferric, zinc and/or nickel chloride. In power delivery (discharging) using iron alone, the reactions are~
At the anode, Fe-~Fe +2e yielding O.44 volts, At the cathode, Fe +e-~Fe yielding O.7~ volts, the overall reaction being:
Fe + 2FeC13--~3FeC12, i.e., 1.2 volts overall.
These reactions are reversible for charging purposes, The oxidizer is recirculated continuously and direct chemical re-action between the oxidizer and solid fuel by physical contact in the fully charged or partially charged system is limited by a microporous membrane barrier between the oxidizer and fuels so that reactions therebetween are predominantly electrochemical The system is a solid/liquid energy converter in which the metal fuel or reducing agent is stored in a solid form within an electrode structure cartridge of the reac-tor The anti-fuel or oxidizer is a liquid which is circulated through the reactor by an electric pump. The electric pump is powered ~: :
by a very small percentage of the electric output of the reactor . .:
^ , .... . .
~'7~3~S~
during opera-tion.
Elec-trodes ~or power produclng cells are within a reactor comprising multiple anoly-te compar-tment and catholyte compartment con-taining cells in elec-trical series or parallel and fluid parallel array, activated only when -the oxidizer is introduced into -the cells. As the system is operated, de-livering power -to an external load, ~the solid fuel is oxidized, solubilized and carried into a recirculating liquid anolyte stream Oxidizer dissolved in pola:r, preferably aqueous, solven-t to form a catholyte is recirculated con-tinuously and such re-circulation involves con-tinuous or semicon~tinuous draw ~rom a reservoir preferably having at least ~ten times -the combined volume of -the catholyte compartments, for maximum energy density of the system, Depending upon the application needs, the reactor can have stored into its structure when charged the equivalent of a single filling or many refillings of oxidizer tanks The addition o~ nickel salt and/or zinc salt in total or partial substitution of iron anolyte salt content reduces the oxidation (rus-ting) vulnerability of iron and reduces polarization. Nickel and zinc a~ford the same anodic multiple valence states as iron and therefore participates in electro-chemical power generation in the same manner as iron ~n re-ducing agents, In applications where oxidation of iron may be a problem9 oxidation inhibitors may be added to the anolyte electrolyte in lieu of nickel usage. These may comprise one or more of citrates, acetate, oxalate and tartarate compounds of alkali metals or ammonium, all of which compounds are soluble in water and alcohol and which readily complex ferrous compounds without affecting eIectrolyte conductivity signifi-cantly A sensor can be applied to anolyte and/or catholyte - 5 ~
~7~35~;) flows -to determine resis-tivity or densi-ty -thereof by optical or elec-trical means and such parame-ters are measures of the s-ta-te of charge (or discharge) of -the sys-tem, The characteristics of the system are as follows, Catholyte is continuously recirculated during operation at a near-constant flow rate independen-t of converter output power level, Since no materials are expelled from the system and no materials are consumed, the total volume of fluid remains essentially the same from the start to the end of a charge, The catholyte reservoir can be removed and replaced by a tank containing new, regenerated fluids, or -the reservoir may be drained of spent catholyte or refilled with fresh ca-tholyte.
When the array has been depleted of i-ts stored fuel, it can be removed for "recharging" and replaced wi-thin a few moments by a freshly "charged" cartridge (or complete system) for con-tinuous device operation, In most applications -the charge or fuel stored in a cartridge will be sufficient for a number of refillings of ozidizer tank fluid.
Some examples of applications of the system are pre-sented below, The weights and sizes given do not include ex-ternal case or pump motors, ~:
(A) Small portable tools and standard electric out-board trolling motors with a 200 to 300 watt range, can be served by a system which provides voltage of 12 volts, current of 15 to 20 amps, and operating times of about 16 hours (4 KWH) for the cartridge and about 2 hours (500 WH) for each catholyte refill to provide averaged power of about 250 watts, ~he siæe of reactor would be 13 lbs, and 0,08 ft3 and the size of the tank would be 15 lbs, (1,16 ft39 i,e. 1.3 gal.), For continuous ~~-operation at 250 watts the fluid must be replenished every two hours and the rechargeable cell array cartridge every 16 hours, (B) Heavy duty industrial -tools, ride-on lawn ' ~793~0 -trac-tors, and larger elec-tric powered boa-ts run at a 2 KW power level These can be accommodated in several modes For in-stance:
Mode 1: Assuming six-teen hours per reactor, two hours per fluid tank, operating times for the cartridge of abou-t 16 hours (32 KWH) and for the oxidizer of about 2 hours (4 KWH) yielding averaged power of about 2 KW, can use a 98 lbs. o.6 f-t3 reac-tor and a 118 lbs, (1.2~ ft39 i.e. 10 gal.) catholyte reservoir tank.
Mode 2: Assuming four hours per reac-tor, -two hours per catho-lyte reservoir, opera-ting times for the cartridge of abou-t four hours (8 KWH) and for the oxidizer of abou-t -two hours (4 KWH) yielding averaged power of about 2 KW can use a 24 6 lbs (0,15 ft3) reactor and a 118 lbs. (1.25 ft3, 10 gal.) reservoir.
(C) A medium sized, 4-passenger, 2500 lb. au-to of the commuter car variety requires short refueling time about 200 WH/mile of energy for propulsion at 40 to 50 mph speeds on level ground and top speed of 70 mph. The presen-t invention can meet these requirements with a system affording: A range of about 1000 miles for the cell array and 100 miles per oxidizer fluid refill; a reactor size of abou-t 650 lbs and~a 4 ft3 volume wi-th 12 KW continuous output capability and about 20 KW peak power; oxidizer fluid tank size of 600 lbs. and 6.3 ft3 volume (50 gal.).
The reactor stores 200 KWH ini-tially and grows pro-gressively ligh-ter by about 9% in weigh-t with each of the ten catholyte reservoir refillings as the car is driven over 1,000 miles The catholyte reservoir holds about 20 KWH of energy per refilling and must be replenished after every 1000 miles of average driving (D) ~oad levelling for power delivery systems and/or emergency or standby power are served through the flexibility ,.~ :
: , :... . . . . ,. ~.. . . . .. .
3~
economy, safe-ty and long li~e charac-teris-tics and versatility of -the energy conversion sys-tem of -the presen-t in-vention, In order to insure long lived opera-tion in any of the ~oregoing applica-tions, the energy conversion system should have means for preventing the formation of, or resolubil-izing, iron oxide precipitate in the anolyte and catholyte.
This is accomplished in accordance wi-th the invention -through certain filtration, alloying, bleed -through an auxiliary energy converter with elec-troly-te :reversal and overpo-ten-tial driving and/or chelation, It is necessary -to address -the ques-tion of pH
variations 'in an iron-redox system -to establish an operable and practical energy source. Iron sal~ts such as the chlorides, sulfates, nitrates, etc. are soluble only in a low pH aqueous medium. Iron oxides, carbonates and hydroxides are liable to formation and precipitation as the pH is allowed -to rise, (usually much above 3), especially where oxygen or other soluble gases such as C02 are available to the solution when exposed -to air.
Ferrous ions in solution are particularly susceptible to oxidation to ferric compounds. Oxygen will reac-t with ferrous salt solutions producing insoluble ferric compounds.
In addition to the above, we have the process of hydrogen gas liberation with which to contend. During the charging process when metallic iron is deposited on the anode surface hydrogen gas is also generated by the decomposition of H20. Even when a cell is on open circui-t some H2 gas is evolved due to the high popula-tion density of-H ions in solu-tions with pH in range of 1 to 2, Unfortunately there normally is not the immediate corresponding evolution of oxygen gas at the cathode to compensate for -the lost hydrogenp with the restoration of ::
pH. As H2 gas evolves, OH ion concentration rises and pH in-,, ~ ' t33rj~3 creases accordingly.
According to the invention, a nurnber o~ conditions are satisfied for the elec-trochemical couple -to be electro-chemically reversible:
~ -use of a single fluid or ionic species in solution which can be separated electrically subsequent to diffusion from one electroly-te to the o-ther.
--use o~ components which can be electrodeposited out of solution in a separable manner.
--means are provided for keeping the primary com-ponents physically separated but available a-t -their respective electrode surfaces for la-ter "discharge".
A review follows of the free energies of formation, (see Oxidation Potentials, 2nd Edition, W.M. ~atimer, Prentice Hall, Inc.), for the various compounds of iron and chlorine which are possible in the fundamental processes associated with the iron-chloride aqueous system.
The formation energies for those compounds of pri-mary concern to the matter of reversibility are listed in -the table below.
~ 9 ~
'~ ' . . ~ , ., ~, . . .~, .. .. . . . . . . .
",, ", ,,",,~ ", ~ ""' ,, ~ "~ ;"""~
33rj~
Free Ener~ of Forma-tion ~a-t 2~C) Substance Ph~sical S-ta-te F K cal~mole H20 ~iq, -56.7 OH aq. -37.6 Fe aq. -20,3 +++
Fe aq. - 2,5 FeC12 solid -72,2 FeC13 solid -80.4 ; FeO solid -58,l~
10 Fe23 solid -177.1 Fe304 solid -242,4 Fe (OH)2 solid -115,6 Fe (OH)3 solid -166,0 Cl aq. -31.~
HCl aq. -31,4 H aq. 00,0 The energy producing reaction results in a net free energy change of Fe + 2Fe ~ 3Fe ~ f + O + 2(-2.5) = 3(-20.3) k cal/mole or ~ F = -55,9 k caljmole, which corresponds to 80 watt-hours per pound of reagents at an open circuit potential of 1,21 volts, -During the charging portion of a cycle some hydrogen - -gas is libera-ted at the anode while plating iron on-to the ; surface. The generation of H2 in this manner tends -to drive the anolyte less acidic causing a precipitation of iron hydroxides and/or oxides, Oxygen dissolved in the elec-trolyte will also tend ~to precipitate the oxides of iron from a ferrous chloride solu-tion. The genera-tion of H2 in this manner -tends to drive the anolyte less acidic causing a precipi-ta-tion of iron hydroxides : , .
.
: ~
: -- 1 0 -- ' 3~
and/or oxides, Oxygen dissolved in -the elec-troly-te will also -tend -to precipita-te -the oxides of iron from a ferrous chloride solution, Inspection of -the energ~ balances for the possible reactions which may result in solubilizing iron oxides and hydroxides shows that the oppor-tunity for structuring an electrically reversible system appears good, When an Fe/FeC13 system i.s being recharged, the Fe ions are being oxidized to -the Fe state a-t the cathode surface, Chloride ions are diffusing to -the ca~thode -to balance the electrical charge, If, however, -the current densi-ty ex-ceeds -the rate with which Fe ions are available for oxidation and if the driving voltage is high enough C12 may be generated a-t the cathode, C12 is moderately soluble in chloride solu-tions with the formation of complex ions, Also C12 will react wi-th water, especially in the presence of active carbon to form HCl.
In any even-t, the production of FeC13 in the cathode results in an increasingly acidic solu-tion with high Cl ion concentration and thelpossibility of free chlorine in small concentrations, Iron oxides are then attacked by the chloride ions and/or free C12 in such a fashion tha-t FeC12 or FeC13 is regenerated with the creation of either H20 or 2 gas as the case may be.
Examination of the following reac-tion possibilities shows that their free energy changes are all in the desired direction. The net free energy change is negative, If HCl is present in sufficient quantities, -then the :
following reactions with the various oxides can be expected as typ:cal proce6ses, ':
, - 11 - '' :
`11~7~
FeO -~ 2H ~Fe + H20 F -58,L~ + o > -20,3 -~6,7 ~ F = -18,6 K cal/mole Fe (OH)2 + 2H --~ Fe + 2H20 F -115,6 + O -~ -20,3 + 2(-56.7) ~ F = -18,1 K ca]/mole Fe (OH)3 -~ 3X Fe + 3H20 F -166 + O~ 2,5 + 3(-'j6,7) ~ F = -6,6 K cal/mole If free chlorine were present in -the electroly-te, some expected reactions would be as follows, ++
FeO + C12 Fe + 2Cl + -2 2 F -~8,4 + O ~-20.3 + 2(-31,4) + O
~ F = -24,7 K cal/mole Fe (OH)2 + C12 ~-~Fe + 2Cl + H20 + 2 2 ,6 + 0--~ -20,3 + 2(-31,4) -56,7 + O
~ F = -24,2 K cal/mole Analysis of the majority of chemical reaction possi-bilities shows that the formation of iron oxides and hydroxides due to electrolysis of water and dissolved oxygen from the air are convertible to the soluble chlorides again by suitable ré-charging, filtering an~ ~luid flow con-trol ~techniques within ;~ the energy converter system, An in-teresting energy consideration is the reaction ; between iron and HCl, Fe and FeC12, Iron will react with HCl to produce the ferrous chloride quite readily, as shown by the following calculations.
~ Fe + 2HCl~ FeC12 + H2 ~;
`~ A
F + O ~ 2(-31,4~ -72,2 -~ O
~F~= -9.4 K cal/mole ; ~ ~ - 12 -:~:
s~
However, -the reac-tlon, ~e -~ 3HCl ~ ~eCl3 -~ 3/2 H2 F -~ 0 -~ 3(-31 L~ - 8 0, L~ -r O -r F
~ F = ~13.8 K cal/mole does no-t wan-t -to -take place as indica~ted by the positive ~ree energy The presence o~ HCl in FeCl2 solu-tion will not result in the oxidation of the ~errous ion, as evidenced by the energy balance below.
FeC12 ~ HCl -~FeCl3 ~~ 2 ~12 F -72.2 - 31 L~ -~ -80.~ ~ 0 ~ F = ~23.2 K cal/mole Any approach -to controlling pH or -the rate o~ accumu-lation of sediment in an aqueous solu-tion FeCl2 and FeCl3 electrolyte secondary cell must be compatible with all o-ther system considerations.
~ he obvious means o~ pH main-tenance by the addition of replacement H ions via HCl is no-t consisten-t with other imperatives since the elec-troly-te is continually driven o~
st~hiometric balance by the addition o~ Cl ions. Eventually there are no ~errous, Fe ~, ions remaining in the system even when in the discharged state due to irreversible oxidation processes. Tests have given about 3% to 5% ~igures for the ampere-hour loss factor or equivalent of H2 gas generation .
during charge. Charging rates are between 6 and 12 amperes for a 60 in2 area electrode, (e.g. abou-t 0.10 to 0 20 amp/in ), ;~ in these experiments. Hence, it would seem on -this basis, -that at the end of 15 to 22 cycles only about 50% o~ the energy storage capacity o~ the electrolyte remains in the ~orm o~
~available Fe i.ons because of the need to continually add HCl to the system. Care must also be taken to charge the system at a correspondingly decreasing ampere-hour value as the cycle 93S~
numbers grow larger otherwise an inordina-te amolm-t o~ H2 will be released by overcharging. The above number~ were arrived a-t by solving for n in -the equa-tions (0,97)n = 0,5, and (0.95)n - 0,5 A second, and more practical me-thod in accordance with the present invention is -to make use of the chemical dynamics at the cathode -to restore pH and solubilize the alkaline pre-cipitates. The catholyte is continuously being driven acidic during the charging process. Filters may be employed to separate ou-t of the fluid stream solid par-ticles which are formed in the anolyte during cycling. Periodically -these fil-ters are switched from the anoly-te to catholyte streams where the collec-ted insoluble iron-compounds are redissolved in the acidic electrolyte. In order to maintain an anoly-te filter in opera-tion at all times two filters are employed and their positions are interchanged periodically.
~ aboratory devices ranging from single cells to 22 cell arrays have been operated in -this fashion over hundreds of cycles with no net, accumulated deterioratlon in performance beyond the first few conditioning cycles. Teal water filter ~. .
cartridges No. lP753 manufactured by the Dayton Elec-tric Manu-facturing ~ompany have been employed successfully over periods of many months of operation. Switching of line filters may be accomplished by two-way valves. They can be switched manually as desired or as -the sediment level and corresponding fluid flow impedence changes would indica-te.
::
Another method which is reasonably effective excep-t .
for the efficiency losses due to electrolyte mixing involves .
the following simple steps, ~ --Place a filter only in the catholyte line.
; ~ ~ --Periodically flush the anolyte through -the catho-; ~ ~ : : .
: :
~117~35~
ly-te by swi-tching the valving such -that the -two hydraulic circui-ts are in series or such tha-t they are bo-th in parallel wi-th the ca-tholyte filter, This periodic flushing of -the anode side with catho-lyte, restores pH and removes precipita-tes which are tending -to collect, Unfor-tunately, the in-troduction of Fe 3 ions into the anolyte reduces efficiency because of the corresponding amp-hour loss by direct attack of iron pla-ting by the ferric solution, We have implemented this me-thod with success, ~osses encountered are proportioned -to fluid volume sizes and frequency of flushing, This loss is reducible by design of a hydraulic circuit which retains most of the anoly-te intact while -the cell is merely flushed out wi-th catholy-te, Alterna-tively, it is possible to have a -two filter system wherein the anolyte fil-ter is reverse flushed periodically into the catho-lyte filter without ever interchanging filter circuits, The residual catholyte in the pores of the anolyte filter after each such operation would also aid in lowering anolyte pH, Partial or to-tal reverse charging of a cell also restores pH and solubilizes iron compounds, The distinct dis-advantage here is -the additional time and energy required to accomplish this end, Even if a cell were to be designed in a symmetrical fashion and operable in either polarity, the energy lost in discharging the cell to zero from the "knee of curve"
and the energy required toraise-the level to the flat character-istic region in the opposite direction are significantly great to cause some dismay for load leveling applications, ~
An "active filter" may be employed in which the filter- :
ing and solubilizing of iron compounds is performed simul-tane-ously at one surface, The ~ilter is an element which is the cathode of asingle cell within a multiple cell energy storage array or as ~: : ~' - -- . .. : ~
~7~S~
a separate auxiliary cell and is cons-tructed of porous carbon.
Composi-te carbons of -this sor-t may comprise ac-tive carbon particles, graphi-te, and PVC binder. The porous carbon filter electrode becomes -the positive electrode of a single eell and -the anolyte is passed -through this electrode filtering out pre-eipita-tes onto the working surface. As electric current is passed through the eell the precipita-tes are resolubilized by beeoming chlorides of iron again by either one of two processes;
the direct electrolytic change from oxides or hydroxides to chlorides at -the electrode surface, typieal half eell reaction Fe23 --~Fe + 2 2 3 If -the potential of this "filter cell" is great enough some free chlorine is genera-ted a-t -the porous ca-thode surfaee and reaetions of this sort will -take plaee, FeO -~ C12 ~Fe + 2Cl + 2 2 Fe(OH)2 + C12 ~Fe -~ 2Cl + H20 ~~ 2 2 by diree-t C12 reaetions with the solid preeipitates Some ehlorine will reae-t with water, (hydrolysis), and HCl, (hydrogen ions), will be generated giving a further deerease in pH in the immediate vieinity of this eleetrode Aetive earbon catalyzes this hydrolysis.
An auxiliary cell may be employed in which 2 gas is evolved and C12 produeed to deerease pH in the anoly-te. The fluid eireuit is essen-tially the same as in the previous method The eathode, (+), eleetrode, fluid line is eonnee-ted into the anolyte eireui-t of a multiple eell array. C12 and HCl is gen-erated in this elee-trolyte by eharging the single eell at a voltage whieh exeeeds the deeomposition po-tential of 1 80 volts.
~; An aetive earbon reae-tion ehamber may be employed to aeeelerate the hydrolysis of C12 Free ehlorine in -the eireuit will tend -to reaet with Fe 2 ions and oxidize them to Fe 3 ions resulting in no effeetive ehange in pH. This - .:
. . .:., : ,. , , , .~, , . -35~
process needs -to be suppressed in favour of hydrolysis wi-th the corresponding genera-tion of H ions, No iron plating is accumulated on -the negative electrode of the auxiliary cell since -the anode fluid is -the catholy-te circuit for the array and usually quite concentrated in oxidizing Fe 3 ions. Hence, -the auxiliary cell may be operated con-tinuously wi-th no need for discharge or recon-ditioning, Opera-ting conditions for the auxiliary cell are adjusted such that -the electrical curren-t flow and fluid flow through the posi-tive elec-trode compar-tmen-t are appropria-te to produce a cell vol-tage over 1,8 volts and H ion generation rate suf'ficient -to adjus-t pH below 3 or L~, As -the fluid flow decreases in the positive compar-tment, -the C12 and H ion generation rate increases for -the same elec-trical current density due to the decreasing availabili-ty of Fe ions, According to the above objects and features, from a broad aspect, the present invention provides an energy con-version apparatus which comprises means defining at leas-t a single electrolytic cell comprising anolyte and catholyte compartments separated by a diff'usion barrier, Means is also provided defining c~node and ca-thode plastic carbon composite electrodes, respectively, in said anolyte and catholyte com- .
partments, Means is further provided def`ining anoly-te and catholyte liquids recirculating in closed loops which contain : the compar-tments respectively, and recircula-ting the bulk s.-torage means for the fluids, the fluids being solutions of :
metal salt which plate out me-tal on the anode during appli- .
cation of charging potential to the electrodes and deplate .
metal and redissolve it into the anolyte during discharge of the system and which change valence states within the catho-lyte without plating, Electrolytic means is also provided ''': .~
33~
for main~taining catholy-te molari-ty be-tween 2,0 and l~,o, Means is also provided for preventing oxlda-tion o~ -the pla-ting/
depla-ting metal to combat polarization and main-tain revers-ibility over extended charge/discharge cycling.
Other objects, fea-tures and advan-tages of the inven-tion will be apparent ~rom the ~ollowing detailed description with reference therein to the accompanying drawing in which:
FIG, 1 is a schematic diagram showing the electrical interconnections of components o~' a por-table power source including an energy converter;
FIG, lA is an exploded view of the mechanical com-ponents o~ such a power source;
FIG, 2 is a sectional view of the energy converter componen-t of the power source;
FIGS, 2A-2D are plan views o~ one o~ the separa-tor elements of the converter;
FIGS, 3 and 4 are schematic diagrams o~ fil-ter crossing and series embodiments of the energy converter of the invention;
FIG, ~ is a schematic diagram of an embodiment o~
the energy conversion system of the invention utilizing an ~ ;
auxiliary cell;
FIGS, 6-1~ are curves of performance data for different fuel cells described in connection with examples 6 .
et seq,; -FIGS, 16A and 16B are schematic cross-sec-tion views ;~ of a cathode portion showing relevant phenomena of -the opera-tion mechanism during charge and discharge processes; and FIGS, 17A and 17D are schematic isome-tric views o~
state of charge sensing systems using colorimetry, DETAI~ED DESCRIPTION OF PREFERRED EMBODIMENTS
Referring now to -the drawing and more particularly to ' ' 335~
FIGS 1 and lA -thereo~, -there is shown an energy converter 10 having -two electroly-te flow circui-ts therein (no-t shown) -an anolyte flow circuit and a ca-tholyte ~low circuit. Anolyte and catholyte are respectively stored in -tanks l] and 12 re-spectively circula-ted by pumps 13 and lL~. The converter pro-vides power a-t output electrodes, viz. anode 15 (negative) and ca-thode 16 (positive) which are respec-tively connected to system output terminals 17 (negative) and 18 (posi-tive). A
main switch is provided for the system at 21 and a star-ting, spring-loaded, switch is provided at 22, a ba-t-tery 23 is used for star-ting the pumps which are powered by elec-tric mo-tors M
and for such s-tarting is momentarily in-terconnected -to such , pumps by depressing switch 22 to put the battery in line with such pumps via protective circuit elements - diodes 24 and 25 and resistor 26, A single pole-double draw relay 29, whose coil 27 is in series with a variable resis-tance 28 connects either of two variable dropping resistors 31 or 32 in-to the pump circuit in response to converter terminal voltage.
A voltmeter is provided across the converter -terminals at 33 and an ammeter is interconnected in series wi-th the con-ver-ter at 34. Fill lines for the anolyte and catholyte cir-cuits are shown at 41 and 42, respectively and return lines therefor are shown at 43 and 44, respectively; wi-th valved parallel drain lines therefor at 45 and 46, respectively, A mechanical arrangement for mounting the system of FIG. 1 as a suitcase-type power pack and comprising a base plate 51 housing the pumps and electrical control and measuring element i5 shown in FIG. 1~. The base plate 51 typically has a width of 23 inches and a depth of 15 inches and a heigh-t of inches and a slanted front containing visual display elements o~ -the meters and control knobs and swi-tches for the circuit ' 3~
elemen-ts 22, 31 and 32 as well as -the main switch 21. The out-pu-t plugs and jacks, 17/18 and 35/36 are in the sides of the base plate. The conver-ter 10, -the tanks 11 and 12 and -the battery 23 are mounted on -top o~ -the base pla-te and are surround-ed by a cowling 52 with a carry handle 53, slots 54 accommodat-ing the drain lines and latching elements 5~ ~or closing up the unit and making the system liftable as a single unit.
Typical dimensions for the -tanks 11, 12 and the con-verter tank 10 and the ba-t-tery are:
tanks: 52'1 wide, 10" long, 9" high and one pound dry weigh-t Converter: 521~ wide, 9" long, 114~ high and 13 pounds dry weigh-t Battery: 2~" wide, 4 3/4" long, ~1" high and 4 pounds dry weight Typical specifica-tions for the hydraulic and electri-cal system are:
Pumps 14 and 1~: Dayton Electric Company Teel, lP811 marine pumps made of (f) Cycolac R~ -material and weighing 0.8 lbs. each including weight of a driving mo-tor therein and opera-ting at 12 volts and 1.5 amperes each, and modified by replacing the stainless s-teel shaf-t with a fiberglass shaf-t;
Battery 23: Beck-Arnley 12N5-3B 4-lb , 12 volts motorcycle s-tarter bat-tery;
Converter 10 Open Circuit Voltage: 24 volts;
Conver-ter 10 Steady State Operation: 25 amperes and 16 volts with an internal resistance of abou-t 3 ohm;
' .
~ * Registered Trademark ;~ ' ~; ... .:
~7~35i~
Tank 12: 1.5 gallons (18 pounds) per hour Tank 11: .7~ gallons; (7 pounds) per hour Rheostat 31: 0 to 5 ohms, 50 wa-tts;
Rheostat 32: 0 to 1 ohm, 10 watts;
Rheostat 28: 0 -to lOk o:hms, Resistor 26: 30 ohms, 5 wa-tts.
Referring now -to FI~. 2, there is shown an energy conver-ter comprising a s-tacked array of uni-t solid-liquid fuel cells, made up of an end anode electrode 110, an end cathode elec-trode 120 and a plurality of bipolar mid-electrodes 130 ` forming first and second unit cells, 140 and 150, respectively at the longitudinal ends of the s-tack, wi-th addi-tional cells 160 formed therebetween by confron-ting bipolar electrodes 130 in electrical series and fluid parallel (i.e. parallel electroly-te feed and parallel electrolyte wi-thdrawal for both anolyte and catholyte)~ -A diffusion barrier separator 145 divides cell 140 into anolyte compartment 141 and ca-tholyte compartment 142.
Similarly, a similar separator 155 divides cell 150 in-to anolyte compartment 151 and catholyte compar-tment 152 and similar separators 165 are provided for the intermediate cells divid-ing them into anoly-te compartmen-ts 161 and catholyte compart-ments 162 While only a small number of cells are shown in the stack for purposes of illustration, i-t will be appreciated that ~: many more cells can be incorporated in such stack on the same principles The end anode electrode 110 comprises a layer 111 of metal which is bonded to a laminated terminal electrode struc- ~
ture comprising conductive ~plas-tic/carbon composite) substra-tes :
- 21 ~
~D'7~35(~1 113, 114 which are sandwiched abou-t an expanded me-tal ~creen Cathode elec-trode 120 comprises conductive sheet 122 with a charcoal coat mounted on a laminated -terminal electrode struc-ture which comprises conductive substrates 123, 124 sandwiched about an expanded me-tal screen 125 Bipolar mid-elec-trodes 130 have metal laye:rs and charcoal layers 132 sandwiched abou-t a central conduc-tive substrate 135 Frame gaskets 143, 144, 153, 15~, 163, 16~ space aid support -the elec-trodes and membranes and comple~te -the enclosure of slab-like electrolyte compar-tments 141, 142, 151, 152, 161, 162, respec-tively, serving as edge walls therefor. The gaske-ts can also provide conduits for feed and return of electroly-te shown in FIG. 2B where gaske-t 143, serving as edge wall for compartment 141 carries feed passage P3 which divides into branch channels 101, 102 to feed anoly-te to upper corners 103, ~ -104 of compartment 141 and re-turn passage P5 which receives anolyte withdrawn from the lower corners of compartment 141 via a similar branched channel arrangemen-t.
The conductive layers 113, 114, 123, 124, 135, and the substrates of cathodes 122, 132 comprise hot pressed mixtures of plastic and carbon, the plastic being polyvinyli-dene fluoride (PVF), polyvinyl chloride, polyethylene or Teflon*, PVF being preferred and the carbon being a finely divided graphite or carbon black The cathode layers 122, 132 comprise charcoa:L pressed into the surface of a carbon plastic substrate which ls in turn bonded to layers 120, 130. The metal layers Illl 131 are hot pressed onto 110, 120, 130 The laminates 113, l:L4, 115 and 123, 124, 125 are separately hot . -~ 30 pressed before adding layers 111 or 112 thereto ~ ' * Registered Trademark : ~ .
~ 935C~
The separa-tors 145, 1~5, :L6~ may be one mil -thick 50~0 porous, elec-trochemical grade polypropylene ~ilm, such as Celgard or 10 mil thick 30% porous Daramic* polyethylene electro-chemical grade ~ilms or other known ion-permeable diffusion-limiting barrier membranes, The end plates 91, 92 and spacer pla-te 93 can be machined or molded to make fluid passages and tie rod holes therein and the marginal edges of the elec-trode, gasket and membrane parts have aligned holes therein. They require appli-cation of sealants prior to assembly, The par-ts are -then assembled as indicated in FIG. 1 and held in compression by tie rods 99 to form a block-like compac-t package, FIGS. 2A-2D are plan views of componen-ts 113, 143, 130 and 93 viewed as shown a-t AA, BB, CC and DD respectively, in FIG, 2, They show the aligned por-ts Pl-P8 running through the marginal portions of those s-tacked components to form longitudinal fluid flow passages. Certain of ports Pl-P8 in end ports are blocked to define longitudinal dead ends of some passages, FIG. 2A shows that sheets 114 and 113 are made of graded resis-tivity so tha-t there is a high resis-tivity between the ports and low resistivity in -the central portion of such sheets where electrode surfaces are bonded, It is important to assure uniformity of elec-troly-te flow through the anolyte and ca-tholyte compartments over the electrode surfaces and to limit short circuit losses between -~
cells, The cell construc-tion minimizes short circui-t losses and enhances uniformi-ty of flow, As shown in FIGS. 2, 2B
~ : ,. .
and 2C wherein a dis-tribution channel 141DC is provided in anolyte compartment ~41 and a distribu-tion channel 142FM is provided in the catholyte compartment 142. These channels * Registered Trademark ':
~'7~350 serve as plenum chambers and are formed in-tegrally wi-th -the compar-tmen-t struc-ture. Similar return channels are provided for the respec-tive compartments in cells 140, 150 and 160, FIG, 2B indica-tes flow pat-tern (arrows F) for anolyte compart-ment 141, The feed passage in P3 branches into lateral delivery channels 101, 102 with respective corner exits 103, 104 at ends of channel 141DC at -the upper corners of compartment 141, Typical cross-sec-tion dimensions for channel 141DC or 142DC are 1/16 X 1/8 inch (or more in the case of 141DC de-pending on the thickness of metal layer employed) and ~the thick-ness of the central portion of compar-tmen-ts 141 and/or 142, established by separation of the anode or ca-thode surfaces from -the confron-ting surface of membrane 145, is typically 1/64 inch. The distribution channel and plenum chamber flow pattern is thus intrinsically established with minimal ex-traneous structure for this purpose and consis-tent wi-th multiple spaced point fluid entry (or drain) at the cell compar-tment edge which allows a significant design length of parallel electrolyte flow paths (the length counterbalancing low electrolyte re-sistivity to establish an acceptably high in-tercell resistance) t~ be establiShed in a small space within the overlapping margin of component cells of the cell stack.
Spacers (indicated for example at 158, 159) are set in the anolyte and ca-tholyte compartments of each cell to es-tablish a uniform spacing of anode and cathode electrodes from the thin barriers 145, 155, 165. The spacers comprise netting with raised crossovers between intersecting synthetic fiber threads (e.g. polypropylene), Alternatively, the net can be pressed into a corrugated form to permit fluid flo~ while maintaining said spacing.
Referring to FI~. 2D, transverse channels D15, D26, D37 and D48 are formed as troughs in plate 93 and they butt . . , ~ .. ~.. .
3~
against subs-tra-te 124 -to ~orm comple-te elongated passages Pl to P5, P2 -to P6, P3 -to P7 and P4 to P8. Passages Pl, P2, P7 and P8 may con-tinue -through pla-te 93 -to ~eed a further array of cells SB2 utilizing common anoly-te and catholy-te recirculating means for both SBl and 2.
- FIG. 2B is a plan view of the spacing gaske-t frame 143 It is made of insulating material, e.g. polyvinyl chloride and is provided with ports Pl-P8 which are punched through during or after frame manufac-ture. If ports Pl-P8 are formed af-ter frame manufac-ture, only a single molding die is necessary to form -the shapes of all frames 143, 144, 163, 164, 123, 124 and the -troughs 101, 102 (upper and lower) therein and a single multiple hole punch can form ports Pl-P8 for all overlapping parts 113, 114, 123, 124, 143, 144, 145, 153, 154, 155, 163, 164, 165. The metal screens 115 and 125 would be short enough to be by-passed by ports Pl-P8 and, in practice, would extend ou-t of electrodes 110 and 120 in a direction ro-tated 90 from the direction shown in FIG 2 for convenience of illustration. -FIG. 2B shows anolyte feed through P3 and drain through P5 and catholyte feed would be in P4 and drain in P6 However, it is useful in some instances to have cross-over, i.e.- anolyte feed at P3, drain at P6 and ca-tholy-te feed at P4, drain at P5, in order -to balance -their flow pa-tterns in the cell compartments by compensating flow resistance in their respective tr1butary feed and drain passages FIG 2C indicates a plan view of the bipolar electrode laminate 130. The raised anode surface 131 is so proportioned in size that its side edges can be readily conformed to sides of frame 144 to prevent side leakage of electrolyte around the electrode.
FIGS 2, 2B and 2C show that the raised elec-trodes, ~ .
~ ~ - . . .
9~5~
e,g., anode layer 131 in bipolar laminate 130 have ~ lesser heigh-t -than -the compar-tmen-ts and are positioned -to form the mani~olds (e.g,, 141DC, 142DC of FIG, 2) .
Typical manufacturing condi~tions which have ~een used ~or making cells of the type shown in FIG, 1 are:
(1,0) for making sheets 114, 124, (1,1) Mix 50~o 3548 Kynar~ (P~F copolymer) wi-th 50%
Dixon Grade 1112 graphite, Sieve PVF copolymer to break up clumps, (1.2) Blend for 3 minu-tes in PK blender wi-thout agi-tator bar plus 3 minutes with agi-ta-tor bar.
(1,3) Sift to break up clumps.
(1,4) Trowel ou-t a slab of this mix-ture 0,100 in.
-thick, However, apply a mix-ture of 60% PVF, 40~o graphite in -the end regions around por-ts Pl-P8 -to achieve resis-tivity graduation as described above in connection with FIG, 2~, (1,5) Preheat for 7 minutes at 350F, (1,6) Press for 7 minutes at 400F at 256 psi, (1.7) Cool for 7 minutes at 50F a-t 256 psi, (1, 8) Release (2,0) ~or making sheets 113, 135 and 123~ -(2,1) Mix 40% PVF copolymer Kynar* grade 3584 with 60~o Dixon Grade 1112 graphite. Sieve copolymer to break up ;; clumps, (2,2) Blend as in (1,2), ; ~ (2,3) Si~t as in (1,3), (2,4) Trowel a slab as in (1,4), (2,5) Preheat as in (1, 5), (2,6 3 Press as in (1,6).
., (2,7) Oool as in (1.7).
(2,8) Release * Registered Trademark ::
~ ~ _ 26 -- ~ .. : :, , , (3.0) for making ca-thodes 132 and 122, (3,1) ~ix:
55% Dixon Grade 1175 graphi-te 5% Dixon Grade XC 72 graphite which has been pulverized to go through a 40 mesh screen 40% PVF copolymer (Kynar* grade 3584) (3,2) Blend (PK blender withou-t agitator bar) 6 minutes (3.3) Sif-t to break up clumps (3.4) Trowel a slab 3/16 in. thick (3,5) Prehea-t at 400F for 7 minu-tes (3, 6) Press a-t 400F, 256 psi for 7 minu-tes (3,7) Cool a-t 50F~ 256 psi for 7 minutes (3,8) Release (3.10) To coat sheets 123, 135 with surface area enhancing layer, (coatings 122, 132) (3,11) Using activated charcoal such as Darco* 20/40, spread a coating 1/8 inch thick on the surface of the sheet (3,12) Preheat the coating and the sheet -to 400F
20 ~or 3 minutes (3.13) Press at 400F for 3 minu-tes at pressure of 200 psi (3,14) Cool at 50 for 3 minutes at a pressure o~
200 psi (3,15) Release.
.
(4,0) To laminate sheets 114 and 113 with copper conductor 115 sheets 123 and I24 with copper conductor 125, (4,1) Sprinkle half complete mo~olayer of mixture of ~; ~ 60~o grade 1112 g:raphite between the copper screen and the 3a ~ two sheets, :~ :
~ ~ * Registered Trademarks .
35~
(4,2) Preheat at 400F for 3 minutes, (4,3) Press a-t 200 psi a-t L~oo F :~or 3 minutes, (4,4) Cool at 50F at 200 psi fo:r 3 minu-tes, The di~fering graphite loading layers 113 and 114 (and of 123 and 124) throughout the major central portions -thereof balance coefficients o~ thermal expansion and produce a flat resultant laminate which is important ~or later assembly and reliable cell dimensioning, (5,0) To make anode sur~aces 111, :L31 sui-table ~or ion plating, (5,1) Trowel a 1/8 inch -thick layer o~ 50 to 200 mesh coconut charcoal onto a conductive shee-t made as in 2,0 or 3,0 above, (5,2) Preheat a-t 400F for 3 minu-tes, (5,3) Press a-t 400F and 200 psi for 3 minu-tes, (5,4) Cool at 50F and 200 psi for 3 minutes, (6,o) To laminate the coated cathode elec-trodes made per (3,0) above, the anode electrodes made per (3,0) above, the assemblies made per (4,0) above and -to the sheets 130 (2,0), (6,1) Assemble layers and prehea-t the assembled parts 350F for 3 minutes, (6,2) Press at 350F for 3 minutes a-t 80 psi, (6,3) Cool under pressure and release~
(6,10) The passages Pl to P8 and tributaries which carry con~ ~
ductive electrolyte need to be insulated as they pass -through ~ -the elec-trodes, This is accomplished by resistivity grading as described above or as follows, (6,11) Mix 95% PVF copolymer (Kynar* 3584~ wi-th 5%
graphite (Dixon Grade 1112), (6,12) Trowel a slab 3/16 -thick, (6,13) Preheat at 400F for 7 minu-tes, (6,14j Press at 400~ for 7 minutes at 135 psi, -Registered Trademark ' 3~
(6.1~) Cool a-t 50F f:or 7 minu-tes at 135 psi, (6,16) Cut into plugs B 1/6~ smaller in dlameter -than the holes in the elec-trodes -themselves, (6.17) Insert plugs B (FIG. 2C) in-to -the holes in the elec-trode shee-t.
(6.18) Preheat electrode sheet and the plug to ~00~
for 7 minutes.
(6,19) Press at 400 F for 7 minu-tes a-t 116 psi.
(6,20) Cool a-t 50F for 7 minutes at 116 psi.
(6.21) Release, This -technique may also be used for elec-trodes 110 and 120 in lieu of resis-tivity grading as described above, (7,0) Fluid passages 101, 102, e-tc,, are cut as grooves into gaske-ts 1~3, 14L~, 153, 154, 163, 164, (8,0) S-tacks of cells are made up and in so doing the mating surfaces are coa-ted with a sealant/adhesive (with coa-tings applied on grooved gasket faces above and below the groove --e,g. above and below groove 101/102 in FIG. 2B) to prevent liquid leakage, The grooved gasket faces butt against the mem-branes 145, 155, 165. Pressure plates (91, 92, 93) are applied at the ends of the stack and tied together with long bolts 99 to compressively grip the cell stack together in a manner well known in the art.
The anolyte and catholyte supplies (anolyte in and catholyte in) are indicated at A,I, and C.I., respectively and the (spent) anolyte and catholy-te drains a-t A,0. and C,0.
respectively (anolyte out, catholyte out).
Therefore, anolyte comes into the stack through longi-tudinal channel P7, and catholyte comes in through longitudinal channel P8 of p3.ate 91. The ~nolyte and catholyte continue in these ~channe]s through SBl to press plate 93 where the . .
anolyte traverses transverse channel 37 to P3 and the catholyte , .
3~
traverses -trans~erse channel D~8 -to P4 The anolyte and ca-tholy-te -travel back longitudinally -through SBl via channels P3 and pl~ respec-tively where -they are dis-tributed to the uni-t cells via ~branch tribu-taries as des-cribed in connec-tion wi-th FIGS. 2B and 2 above. Af-ter passing through the unit cells, and genera-ting elec-trical power -there, the anolyte and catholyte emerge, via branched -tributaries, to respective return pa-ths P~ and P6, then -traverse transverse channels 26 and 15, respectively, then respectively traverse longitudinal paths Pl and P2, an.d then re-emerge from the cell stack on -the ou-tside of plate 91, as indica-ted by -the lines A 0. and CØ
Separator screens (not shown) are included in each of the anolyte and catholyte compar~tmen-ts to preven-t the flexible diffusion barrier from deflecting over into contact with other elec-trodes confronting it Suitable screens include Dupont's Vexar* Model 10 PDS 169, which has high density poly-ethylene strands of 10 mil width criss-crossing in a diamond pattern with 16 strands per inch and affording 90~0 porosity and which can be corrugated -to various dep-ths of corrugation for fitting in different volumes of Hercules' Delnet* brand embossed and stretched polypropylene sheet, Model GQ 330, affording a 75% porosity and having a thickness o~ 10-12 mils and raised but-ton embossed projections to any desired separation :.
depth The anolyte and catholyte compartments may be di~er-entially pressurized to establish a bias agains-t diffusion of undesired ionic species. Such differential pressurization can :
be achieved through different respective pump sizes and/or :
~ -30 speeds, flow passage sizes; or through throttle valves. ~ :
: ' * Registered Trademark .
11 ~37935i~
An alterna-tive or supplemen-t -to pressure differential diffusion suppression is -the use of a mi~-compartment (not shown) in each uni-t fuel cell between anolyte and catholy-te compartments and separated -therefrom by diffusion barriers or coarser colloid barriers. Such mid-compartments, if used, may be provided with their own recirculating sys-tems with filters -to trap certain diffusing species or may be pressurized to suppress in ei-ther direction.
Internal resistances in the converter may comprise the following components:
resistivity of anolyte: e.g. 2-6 ohm-in. for initial 20%
NaCl in water average over operation of a converter resistivity of catholyte: e.g. 4-10 ohm-in. for 4 Molar FeC13 in water average over operation of a converter resistance of 60 sq in. separator: about.O5 ohm-in. for .010"
thick 30% DARAMIC saturated with 25% NaCl solution or equivalent Other losses are pumping requirements to move electrolytes. In a typical case for 60 sq. in. electrodes at 0.5 amp/sq. in. current drain the pumps move anolyte at O -to 1 cc/sec./cell, catholyte at 1 to 2 cc/sec./cell under a 8 to 10 foot head with a pressure differential between anolyte and -catholyte corresponding to 1-2 feet of head.
~ Spacing in the electrolyte compartments for purposes ; of resistance calculation chanees during a discharge. Anode to barrier spacing grows from .015 inches as metal deposited earlier during charge is consumed FLO~ RATE
BACKGROUND o~i T~IE INVENTION
The present inventlon re:La-tes to energy conversion and more par-ticularl~ to a rechargeable and/or refuela~le high specific power delivery and high specific energy s-torage sys-tem A large number of por-table power delivery and energy storage electrical devices inc]uding automobiles, golf car-ts, forklift trucks, garden appliances, cons-truction equipmen-t, portable motors and lighting uni-ts, advan-tage of Fe/~e 3 here require long life and reliable means for high specific power delivery and high specific energy s-torage capability ("specific"
meaning with respec-t -to weigh-t). ~ow cost and long life energy storage systems are needed in -the above applications and also in connecti.on with data processors and instrumen-ts and load levelling The state of the art, including the most promising of current R&D efforts, is summarized in Iammartino, "New Batteries Are Coming", Chemical Engineering magazine, pp ~8-50, January 20, 1975, The research and development effor-ts include use of the electrochemical pairs zinc-nickel oxide, metal-air (oxygen), sodium-sulfur, li-thium-sulfide, zinc-chlorine and my own prior efforts with iron-ferric chloride.
It is an important object of the present invention to provide such a system free of -the disadvantages of prior systems I-t is a further object of the present invention to provide low cos-t of initial capital equipmen-t consistent with ;~; the preceding object, ; It is a further object of the invention to provide low operating costs consistent with one or more of the pre- ~ -ceding objects.
It is a further object o~ the invention to provide indefinitely long life of capital equipment consistent with ; one or more of the preceding objects.
; ~ ':' :, -- 2 -- .
'793~i~
I-t is a fur-ther objec-t of -the invention to provide subs-tan-tially hazard-free opera-tion and handling consistent with one or more of the preceding objec-ts.
It is a further object of -the inven-tion to provide capabili-ty of being "refueled" immediately consistent with one or more of the preceding objects.
It is a further object of the inven-tion to provide capability of being recharged electrically consis-tent wi-th one or more of the preceding objects.
It is a further object of the invention -to avoid -the necessity for a complex chemical supply network or logis-tics consistent with one or more of the preceding objects.
It is a further object of the inven-tion -to provide odorless operation consistent with one or more of the preceding objects.
It is a further object of the invention to provide low freezing temperatures consistent with one or more of the preceding objects, It is a further objec-t of the invention to provide for complete discharge wi-th no damage consistent with one or more of the preceding objects.
It is a further object of the invention -to provide simple means of indicating remaining energy consistent with one or more of the preceding objects.
It is a further object of the invention to provide -indefinitely long charge retention consis-ten-t with one or more of the preceding objects.
It is a~further object of -the inven-tion to provide that when the system is turned off, there is no power available at the terminals, thus making the system electrically safe to handle when "deactivated" conslstent wi-th one or more of the preceding objects, - ;~
It ls a further object of -the inven-tion -to provide ex-tremely plen-tiful and non-pollu-ting ma-terials ernployed con-sistent with one or more of -the preceding objects.
It is a further objec~ of the inven-tion -to provide an energy conversion device ~or which simple manu~acturing methods can be employed consisten-t with one or more of -the pre-ceding objec-ts.
SUMMARY OF THE INVENTION
According to the inven-tion, an energy conversion system is made in a fashion to be rechargeable in si-tu as a secondary battery or, alternatively, ins-tan-tly refuelable in the fashion of a solid li.quid fuel cell and comprises a ca-thode and, when charged, a solid anode fuel comprising iron, zinc and/or nickel and a liquid oxidizer comprising ferric, zinc and/or nickel chloride. In power delivery (discharging) using iron alone, the reactions are~
At the anode, Fe-~Fe +2e yielding O.44 volts, At the cathode, Fe +e-~Fe yielding O.7~ volts, the overall reaction being:
Fe + 2FeC13--~3FeC12, i.e., 1.2 volts overall.
These reactions are reversible for charging purposes, The oxidizer is recirculated continuously and direct chemical re-action between the oxidizer and solid fuel by physical contact in the fully charged or partially charged system is limited by a microporous membrane barrier between the oxidizer and fuels so that reactions therebetween are predominantly electrochemical The system is a solid/liquid energy converter in which the metal fuel or reducing agent is stored in a solid form within an electrode structure cartridge of the reac-tor The anti-fuel or oxidizer is a liquid which is circulated through the reactor by an electric pump. The electric pump is powered ~: :
by a very small percentage of the electric output of the reactor . .:
^ , .... . .
~'7~3~S~
during opera-tion.
Elec-trodes ~or power produclng cells are within a reactor comprising multiple anoly-te compar-tment and catholyte compartment con-taining cells in elec-trical series or parallel and fluid parallel array, activated only when -the oxidizer is introduced into -the cells. As the system is operated, de-livering power -to an external load, ~the solid fuel is oxidized, solubilized and carried into a recirculating liquid anolyte stream Oxidizer dissolved in pola:r, preferably aqueous, solven-t to form a catholyte is recirculated con-tinuously and such re-circulation involves con-tinuous or semicon~tinuous draw ~rom a reservoir preferably having at least ~ten times -the combined volume of -the catholyte compartments, for maximum energy density of the system, Depending upon the application needs, the reactor can have stored into its structure when charged the equivalent of a single filling or many refillings of oxidizer tanks The addition o~ nickel salt and/or zinc salt in total or partial substitution of iron anolyte salt content reduces the oxidation (rus-ting) vulnerability of iron and reduces polarization. Nickel and zinc a~ford the same anodic multiple valence states as iron and therefore participates in electro-chemical power generation in the same manner as iron ~n re-ducing agents, In applications where oxidation of iron may be a problem9 oxidation inhibitors may be added to the anolyte electrolyte in lieu of nickel usage. These may comprise one or more of citrates, acetate, oxalate and tartarate compounds of alkali metals or ammonium, all of which compounds are soluble in water and alcohol and which readily complex ferrous compounds without affecting eIectrolyte conductivity signifi-cantly A sensor can be applied to anolyte and/or catholyte - 5 ~
~7~35~;) flows -to determine resis-tivity or densi-ty -thereof by optical or elec-trical means and such parame-ters are measures of the s-ta-te of charge (or discharge) of -the sys-tem, The characteristics of the system are as follows, Catholyte is continuously recirculated during operation at a near-constant flow rate independen-t of converter output power level, Since no materials are expelled from the system and no materials are consumed, the total volume of fluid remains essentially the same from the start to the end of a charge, The catholyte reservoir can be removed and replaced by a tank containing new, regenerated fluids, or -the reservoir may be drained of spent catholyte or refilled with fresh ca-tholyte.
When the array has been depleted of i-ts stored fuel, it can be removed for "recharging" and replaced wi-thin a few moments by a freshly "charged" cartridge (or complete system) for con-tinuous device operation, In most applications -the charge or fuel stored in a cartridge will be sufficient for a number of refillings of ozidizer tank fluid.
Some examples of applications of the system are pre-sented below, The weights and sizes given do not include ex-ternal case or pump motors, ~:
(A) Small portable tools and standard electric out-board trolling motors with a 200 to 300 watt range, can be served by a system which provides voltage of 12 volts, current of 15 to 20 amps, and operating times of about 16 hours (4 KWH) for the cartridge and about 2 hours (500 WH) for each catholyte refill to provide averaged power of about 250 watts, ~he siæe of reactor would be 13 lbs, and 0,08 ft3 and the size of the tank would be 15 lbs, (1,16 ft39 i,e. 1.3 gal.), For continuous ~~-operation at 250 watts the fluid must be replenished every two hours and the rechargeable cell array cartridge every 16 hours, (B) Heavy duty industrial -tools, ride-on lawn ' ~793~0 -trac-tors, and larger elec-tric powered boa-ts run at a 2 KW power level These can be accommodated in several modes For in-stance:
Mode 1: Assuming six-teen hours per reactor, two hours per fluid tank, operating times for the cartridge of abou-t 16 hours (32 KWH) and for the oxidizer of about 2 hours (4 KWH) yielding averaged power of about 2 KW, can use a 98 lbs. o.6 f-t3 reac-tor and a 118 lbs, (1.2~ ft39 i.e. 10 gal.) catholyte reservoir tank.
Mode 2: Assuming four hours per reac-tor, -two hours per catho-lyte reservoir, opera-ting times for the cartridge of abou-t four hours (8 KWH) and for the oxidizer of abou-t -two hours (4 KWH) yielding averaged power of about 2 KW can use a 24 6 lbs (0,15 ft3) reactor and a 118 lbs. (1.25 ft3, 10 gal.) reservoir.
(C) A medium sized, 4-passenger, 2500 lb. au-to of the commuter car variety requires short refueling time about 200 WH/mile of energy for propulsion at 40 to 50 mph speeds on level ground and top speed of 70 mph. The presen-t invention can meet these requirements with a system affording: A range of about 1000 miles for the cell array and 100 miles per oxidizer fluid refill; a reactor size of abou-t 650 lbs and~a 4 ft3 volume wi-th 12 KW continuous output capability and about 20 KW peak power; oxidizer fluid tank size of 600 lbs. and 6.3 ft3 volume (50 gal.).
The reactor stores 200 KWH ini-tially and grows pro-gressively ligh-ter by about 9% in weigh-t with each of the ten catholyte reservoir refillings as the car is driven over 1,000 miles The catholyte reservoir holds about 20 KWH of energy per refilling and must be replenished after every 1000 miles of average driving (D) ~oad levelling for power delivery systems and/or emergency or standby power are served through the flexibility ,.~ :
: , :... . . . . ,. ~.. . . . .. .
3~
economy, safe-ty and long li~e charac-teris-tics and versatility of -the energy conversion sys-tem of -the presen-t in-vention, In order to insure long lived opera-tion in any of the ~oregoing applica-tions, the energy conversion system should have means for preventing the formation of, or resolubil-izing, iron oxide precipitate in the anolyte and catholyte.
This is accomplished in accordance wi-th the invention -through certain filtration, alloying, bleed -through an auxiliary energy converter with elec-troly-te :reversal and overpo-ten-tial driving and/or chelation, It is necessary -to address -the ques-tion of pH
variations 'in an iron-redox system -to establish an operable and practical energy source. Iron sal~ts such as the chlorides, sulfates, nitrates, etc. are soluble only in a low pH aqueous medium. Iron oxides, carbonates and hydroxides are liable to formation and precipitation as the pH is allowed -to rise, (usually much above 3), especially where oxygen or other soluble gases such as C02 are available to the solution when exposed -to air.
Ferrous ions in solution are particularly susceptible to oxidation to ferric compounds. Oxygen will reac-t with ferrous salt solutions producing insoluble ferric compounds.
In addition to the above, we have the process of hydrogen gas liberation with which to contend. During the charging process when metallic iron is deposited on the anode surface hydrogen gas is also generated by the decomposition of H20. Even when a cell is on open circui-t some H2 gas is evolved due to the high popula-tion density of-H ions in solu-tions with pH in range of 1 to 2, Unfortunately there normally is not the immediate corresponding evolution of oxygen gas at the cathode to compensate for -the lost hydrogenp with the restoration of ::
pH. As H2 gas evolves, OH ion concentration rises and pH in-,, ~ ' t33rj~3 creases accordingly.
According to the invention, a nurnber o~ conditions are satisfied for the elec-trochemical couple -to be electro-chemically reversible:
~ -use of a single fluid or ionic species in solution which can be separated electrically subsequent to diffusion from one electroly-te to the o-ther.
--use o~ components which can be electrodeposited out of solution in a separable manner.
--means are provided for keeping the primary com-ponents physically separated but available a-t -their respective electrode surfaces for la-ter "discharge".
A review follows of the free energies of formation, (see Oxidation Potentials, 2nd Edition, W.M. ~atimer, Prentice Hall, Inc.), for the various compounds of iron and chlorine which are possible in the fundamental processes associated with the iron-chloride aqueous system.
The formation energies for those compounds of pri-mary concern to the matter of reversibility are listed in -the table below.
~ 9 ~
'~ ' . . ~ , ., ~, . . .~, .. .. . . . . . . .
",, ", ,,",,~ ", ~ ""' ,, ~ "~ ;"""~
33rj~
Free Ener~ of Forma-tion ~a-t 2~C) Substance Ph~sical S-ta-te F K cal~mole H20 ~iq, -56.7 OH aq. -37.6 Fe aq. -20,3 +++
Fe aq. - 2,5 FeC12 solid -72,2 FeC13 solid -80.4 ; FeO solid -58,l~
10 Fe23 solid -177.1 Fe304 solid -242,4 Fe (OH)2 solid -115,6 Fe (OH)3 solid -166,0 Cl aq. -31.~
HCl aq. -31,4 H aq. 00,0 The energy producing reaction results in a net free energy change of Fe + 2Fe ~ 3Fe ~ f + O + 2(-2.5) = 3(-20.3) k cal/mole or ~ F = -55,9 k caljmole, which corresponds to 80 watt-hours per pound of reagents at an open circuit potential of 1,21 volts, -During the charging portion of a cycle some hydrogen - -gas is libera-ted at the anode while plating iron on-to the ; surface. The generation of H2 in this manner tends -to drive the anolyte less acidic causing a precipitation of iron hydroxides and/or oxides, Oxygen dissolved in the elec-trolyte will also tend ~to precipitate the oxides of iron from a ferrous chloride solu-tion. The genera-tion of H2 in this manner -tends to drive the anolyte less acidic causing a precipi-ta-tion of iron hydroxides : , .
.
: ~
: -- 1 0 -- ' 3~
and/or oxides, Oxygen dissolved in -the elec-troly-te will also -tend -to precipita-te -the oxides of iron from a ferrous chloride solution, Inspection of -the energ~ balances for the possible reactions which may result in solubilizing iron oxides and hydroxides shows that the oppor-tunity for structuring an electrically reversible system appears good, When an Fe/FeC13 system i.s being recharged, the Fe ions are being oxidized to -the Fe state a-t the cathode surface, Chloride ions are diffusing to -the ca~thode -to balance the electrical charge, If, however, -the current densi-ty ex-ceeds -the rate with which Fe ions are available for oxidation and if the driving voltage is high enough C12 may be generated a-t the cathode, C12 is moderately soluble in chloride solu-tions with the formation of complex ions, Also C12 will react wi-th water, especially in the presence of active carbon to form HCl.
In any even-t, the production of FeC13 in the cathode results in an increasingly acidic solu-tion with high Cl ion concentration and thelpossibility of free chlorine in small concentrations, Iron oxides are then attacked by the chloride ions and/or free C12 in such a fashion tha-t FeC12 or FeC13 is regenerated with the creation of either H20 or 2 gas as the case may be.
Examination of the following reac-tion possibilities shows that their free energy changes are all in the desired direction. The net free energy change is negative, If HCl is present in sufficient quantities, -then the :
following reactions with the various oxides can be expected as typ:cal proce6ses, ':
, - 11 - '' :
`11~7~
FeO -~ 2H ~Fe + H20 F -58,L~ + o > -20,3 -~6,7 ~ F = -18,6 K cal/mole Fe (OH)2 + 2H --~ Fe + 2H20 F -115,6 + O -~ -20,3 + 2(-56.7) ~ F = -18,1 K ca]/mole Fe (OH)3 -~ 3X Fe + 3H20 F -166 + O~ 2,5 + 3(-'j6,7) ~ F = -6,6 K cal/mole If free chlorine were present in -the electroly-te, some expected reactions would be as follows, ++
FeO + C12 Fe + 2Cl + -2 2 F -~8,4 + O ~-20.3 + 2(-31,4) + O
~ F = -24,7 K cal/mole Fe (OH)2 + C12 ~-~Fe + 2Cl + H20 + 2 2 ,6 + 0--~ -20,3 + 2(-31,4) -56,7 + O
~ F = -24,2 K cal/mole Analysis of the majority of chemical reaction possi-bilities shows that the formation of iron oxides and hydroxides due to electrolysis of water and dissolved oxygen from the air are convertible to the soluble chlorides again by suitable ré-charging, filtering an~ ~luid flow con-trol ~techniques within ;~ the energy converter system, An in-teresting energy consideration is the reaction ; between iron and HCl, Fe and FeC12, Iron will react with HCl to produce the ferrous chloride quite readily, as shown by the following calculations.
~ Fe + 2HCl~ FeC12 + H2 ~;
`~ A
F + O ~ 2(-31,4~ -72,2 -~ O
~F~= -9.4 K cal/mole ; ~ ~ - 12 -:~:
s~
However, -the reac-tlon, ~e -~ 3HCl ~ ~eCl3 -~ 3/2 H2 F -~ 0 -~ 3(-31 L~ - 8 0, L~ -r O -r F
~ F = ~13.8 K cal/mole does no-t wan-t -to -take place as indica~ted by the positive ~ree energy The presence o~ HCl in FeCl2 solu-tion will not result in the oxidation of the ~errous ion, as evidenced by the energy balance below.
FeC12 ~ HCl -~FeCl3 ~~ 2 ~12 F -72.2 - 31 L~ -~ -80.~ ~ 0 ~ F = ~23.2 K cal/mole Any approach -to controlling pH or -the rate o~ accumu-lation of sediment in an aqueous solu-tion FeCl2 and FeCl3 electrolyte secondary cell must be compatible with all o-ther system considerations.
~ he obvious means o~ pH main-tenance by the addition of replacement H ions via HCl is no-t consisten-t with other imperatives since the elec-troly-te is continually driven o~
st~hiometric balance by the addition o~ Cl ions. Eventually there are no ~errous, Fe ~, ions remaining in the system even when in the discharged state due to irreversible oxidation processes. Tests have given about 3% to 5% ~igures for the ampere-hour loss factor or equivalent of H2 gas generation .
during charge. Charging rates are between 6 and 12 amperes for a 60 in2 area electrode, (e.g. abou-t 0.10 to 0 20 amp/in ), ;~ in these experiments. Hence, it would seem on -this basis, -that at the end of 15 to 22 cycles only about 50% o~ the energy storage capacity o~ the electrolyte remains in the ~orm o~
~available Fe i.ons because of the need to continually add HCl to the system. Care must also be taken to charge the system at a correspondingly decreasing ampere-hour value as the cycle 93S~
numbers grow larger otherwise an inordina-te amolm-t o~ H2 will be released by overcharging. The above number~ were arrived a-t by solving for n in -the equa-tions (0,97)n = 0,5, and (0.95)n - 0,5 A second, and more practical me-thod in accordance with the present invention is -to make use of the chemical dynamics at the cathode -to restore pH and solubilize the alkaline pre-cipitates. The catholyte is continuously being driven acidic during the charging process. Filters may be employed to separate ou-t of the fluid stream solid par-ticles which are formed in the anolyte during cycling. Periodically -these fil-ters are switched from the anoly-te to catholyte streams where the collec-ted insoluble iron-compounds are redissolved in the acidic electrolyte. In order to maintain an anoly-te filter in opera-tion at all times two filters are employed and their positions are interchanged periodically.
~ aboratory devices ranging from single cells to 22 cell arrays have been operated in -this fashion over hundreds of cycles with no net, accumulated deterioratlon in performance beyond the first few conditioning cycles. Teal water filter ~. .
cartridges No. lP753 manufactured by the Dayton Elec-tric Manu-facturing ~ompany have been employed successfully over periods of many months of operation. Switching of line filters may be accomplished by two-way valves. They can be switched manually as desired or as -the sediment level and corresponding fluid flow impedence changes would indica-te.
::
Another method which is reasonably effective excep-t .
for the efficiency losses due to electrolyte mixing involves .
the following simple steps, ~ --Place a filter only in the catholyte line.
; ~ ~ --Periodically flush the anolyte through -the catho-; ~ ~ : : .
: :
~117~35~
ly-te by swi-tching the valving such -that the -two hydraulic circui-ts are in series or such tha-t they are bo-th in parallel wi-th the ca-tholyte filter, This periodic flushing of -the anode side with catho-lyte, restores pH and removes precipita-tes which are tending -to collect, Unfor-tunately, the in-troduction of Fe 3 ions into the anolyte reduces efficiency because of the corresponding amp-hour loss by direct attack of iron pla-ting by the ferric solution, We have implemented this me-thod with success, ~osses encountered are proportioned -to fluid volume sizes and frequency of flushing, This loss is reducible by design of a hydraulic circuit which retains most of the anoly-te intact while -the cell is merely flushed out wi-th catholy-te, Alterna-tively, it is possible to have a -two filter system wherein the anolyte fil-ter is reverse flushed periodically into the catho-lyte filter without ever interchanging filter circuits, The residual catholyte in the pores of the anolyte filter after each such operation would also aid in lowering anolyte pH, Partial or to-tal reverse charging of a cell also restores pH and solubilizes iron compounds, The distinct dis-advantage here is -the additional time and energy required to accomplish this end, Even if a cell were to be designed in a symmetrical fashion and operable in either polarity, the energy lost in discharging the cell to zero from the "knee of curve"
and the energy required toraise-the level to the flat character-istic region in the opposite direction are significantly great to cause some dismay for load leveling applications, ~
An "active filter" may be employed in which the filter- :
ing and solubilizing of iron compounds is performed simul-tane-ously at one surface, The ~ilter is an element which is the cathode of asingle cell within a multiple cell energy storage array or as ~: : ~' - -- . .. : ~
~7~S~
a separate auxiliary cell and is cons-tructed of porous carbon.
Composi-te carbons of -this sor-t may comprise ac-tive carbon particles, graphi-te, and PVC binder. The porous carbon filter electrode becomes -the positive electrode of a single eell and -the anolyte is passed -through this electrode filtering out pre-eipita-tes onto the working surface. As electric current is passed through the eell the precipita-tes are resolubilized by beeoming chlorides of iron again by either one of two processes;
the direct electrolytic change from oxides or hydroxides to chlorides at -the electrode surface, typieal half eell reaction Fe23 --~Fe + 2 2 3 If -the potential of this "filter cell" is great enough some free chlorine is genera-ted a-t -the porous ca-thode surfaee and reaetions of this sort will -take plaee, FeO -~ C12 ~Fe + 2Cl + 2 2 Fe(OH)2 + C12 ~Fe -~ 2Cl + H20 ~~ 2 2 by diree-t C12 reaetions with the solid preeipitates Some ehlorine will reae-t with water, (hydrolysis), and HCl, (hydrogen ions), will be generated giving a further deerease in pH in the immediate vieinity of this eleetrode Aetive earbon catalyzes this hydrolysis.
An auxiliary cell may be employed in which 2 gas is evolved and C12 produeed to deerease pH in the anoly-te. The fluid eireuit is essen-tially the same as in the previous method The eathode, (+), eleetrode, fluid line is eonnee-ted into the anolyte eireui-t of a multiple eell array. C12 and HCl is gen-erated in this elee-trolyte by eharging the single eell at a voltage whieh exeeeds the deeomposition po-tential of 1 80 volts.
~; An aetive earbon reae-tion ehamber may be employed to aeeelerate the hydrolysis of C12 Free ehlorine in -the eireuit will tend -to reaet with Fe 2 ions and oxidize them to Fe 3 ions resulting in no effeetive ehange in pH. This - .:
. . .:., : ,. , , , .~, , . -35~
process needs -to be suppressed in favour of hydrolysis wi-th the corresponding genera-tion of H ions, No iron plating is accumulated on -the negative electrode of the auxiliary cell since -the anode fluid is -the catholy-te circuit for the array and usually quite concentrated in oxidizing Fe 3 ions. Hence, -the auxiliary cell may be operated con-tinuously wi-th no need for discharge or recon-ditioning, Opera-ting conditions for the auxiliary cell are adjusted such that -the electrical curren-t flow and fluid flow through the posi-tive elec-trode compar-tmen-t are appropria-te to produce a cell vol-tage over 1,8 volts and H ion generation rate suf'ficient -to adjus-t pH below 3 or L~, As -the fluid flow decreases in the positive compar-tment, -the C12 and H ion generation rate increases for -the same elec-trical current density due to the decreasing availabili-ty of Fe ions, According to the above objects and features, from a broad aspect, the present invention provides an energy con-version apparatus which comprises means defining at leas-t a single electrolytic cell comprising anolyte and catholyte compartments separated by a diff'usion barrier, Means is also provided defining c~node and ca-thode plastic carbon composite electrodes, respectively, in said anolyte and catholyte com- .
partments, Means is further provided def`ining anoly-te and catholyte liquids recirculating in closed loops which contain : the compar-tments respectively, and recircula-ting the bulk s.-torage means for the fluids, the fluids being solutions of :
metal salt which plate out me-tal on the anode during appli- .
cation of charging potential to the electrodes and deplate .
metal and redissolve it into the anolyte during discharge of the system and which change valence states within the catho-lyte without plating, Electrolytic means is also provided ''': .~
33~
for main~taining catholy-te molari-ty be-tween 2,0 and l~,o, Means is also provided for preventing oxlda-tion o~ -the pla-ting/
depla-ting metal to combat polarization and main-tain revers-ibility over extended charge/discharge cycling.
Other objects, fea-tures and advan-tages of the inven-tion will be apparent ~rom the ~ollowing detailed description with reference therein to the accompanying drawing in which:
FIG, 1 is a schematic diagram showing the electrical interconnections of components o~' a por-table power source including an energy converter;
FIG, lA is an exploded view of the mechanical com-ponents o~ such a power source;
FIG, 2 is a sectional view of the energy converter componen-t of the power source;
FIGS, 2A-2D are plan views o~ one o~ the separa-tor elements of the converter;
FIGS, 3 and 4 are schematic diagrams o~ fil-ter crossing and series embodiments of the energy converter of the invention;
FIG, ~ is a schematic diagram of an embodiment o~
the energy conversion system of the invention utilizing an ~ ;
auxiliary cell;
FIGS, 6-1~ are curves of performance data for different fuel cells described in connection with examples 6 .
et seq,; -FIGS, 16A and 16B are schematic cross-sec-tion views ;~ of a cathode portion showing relevant phenomena of -the opera-tion mechanism during charge and discharge processes; and FIGS, 17A and 17D are schematic isome-tric views o~
state of charge sensing systems using colorimetry, DETAI~ED DESCRIPTION OF PREFERRED EMBODIMENTS
Referring now to -the drawing and more particularly to ' ' 335~
FIGS 1 and lA -thereo~, -there is shown an energy converter 10 having -two electroly-te flow circui-ts therein (no-t shown) -an anolyte flow circuit and a ca-tholyte ~low circuit. Anolyte and catholyte are respectively stored in -tanks l] and 12 re-spectively circula-ted by pumps 13 and lL~. The converter pro-vides power a-t output electrodes, viz. anode 15 (negative) and ca-thode 16 (positive) which are respec-tively connected to system output terminals 17 (negative) and 18 (posi-tive). A
main switch is provided for the system at 21 and a star-ting, spring-loaded, switch is provided at 22, a ba-t-tery 23 is used for star-ting the pumps which are powered by elec-tric mo-tors M
and for such s-tarting is momentarily in-terconnected -to such , pumps by depressing switch 22 to put the battery in line with such pumps via protective circuit elements - diodes 24 and 25 and resistor 26, A single pole-double draw relay 29, whose coil 27 is in series with a variable resis-tance 28 connects either of two variable dropping resistors 31 or 32 in-to the pump circuit in response to converter terminal voltage.
A voltmeter is provided across the converter -terminals at 33 and an ammeter is interconnected in series wi-th the con-ver-ter at 34. Fill lines for the anolyte and catholyte cir-cuits are shown at 41 and 42, respectively and return lines therefor are shown at 43 and 44, respectively; wi-th valved parallel drain lines therefor at 45 and 46, respectively, A mechanical arrangement for mounting the system of FIG. 1 as a suitcase-type power pack and comprising a base plate 51 housing the pumps and electrical control and measuring element i5 shown in FIG. 1~. The base plate 51 typically has a width of 23 inches and a depth of 15 inches and a heigh-t of inches and a slanted front containing visual display elements o~ -the meters and control knobs and swi-tches for the circuit ' 3~
elemen-ts 22, 31 and 32 as well as -the main switch 21. The out-pu-t plugs and jacks, 17/18 and 35/36 are in the sides of the base plate. The conver-ter 10, -the tanks 11 and 12 and -the battery 23 are mounted on -top o~ -the base pla-te and are surround-ed by a cowling 52 with a carry handle 53, slots 54 accommodat-ing the drain lines and latching elements 5~ ~or closing up the unit and making the system liftable as a single unit.
Typical dimensions for the -tanks 11, 12 and the con-verter tank 10 and the ba-t-tery are:
tanks: 52'1 wide, 10" long, 9" high and one pound dry weigh-t Converter: 521~ wide, 9" long, 114~ high and 13 pounds dry weigh-t Battery: 2~" wide, 4 3/4" long, ~1" high and 4 pounds dry weight Typical specifica-tions for the hydraulic and electri-cal system are:
Pumps 14 and 1~: Dayton Electric Company Teel, lP811 marine pumps made of (f) Cycolac R~ -material and weighing 0.8 lbs. each including weight of a driving mo-tor therein and opera-ting at 12 volts and 1.5 amperes each, and modified by replacing the stainless s-teel shaf-t with a fiberglass shaf-t;
Battery 23: Beck-Arnley 12N5-3B 4-lb , 12 volts motorcycle s-tarter bat-tery;
Converter 10 Open Circuit Voltage: 24 volts;
Conver-ter 10 Steady State Operation: 25 amperes and 16 volts with an internal resistance of abou-t 3 ohm;
' .
~ * Registered Trademark ;~ ' ~; ... .:
~7~35i~
Tank 12: 1.5 gallons (18 pounds) per hour Tank 11: .7~ gallons; (7 pounds) per hour Rheostat 31: 0 to 5 ohms, 50 wa-tts;
Rheostat 32: 0 to 1 ohm, 10 watts;
Rheostat 28: 0 -to lOk o:hms, Resistor 26: 30 ohms, 5 wa-tts.
Referring now -to FI~. 2, there is shown an energy conver-ter comprising a s-tacked array of uni-t solid-liquid fuel cells, made up of an end anode electrode 110, an end cathode elec-trode 120 and a plurality of bipolar mid-electrodes 130 ` forming first and second unit cells, 140 and 150, respectively at the longitudinal ends of the s-tack, wi-th addi-tional cells 160 formed therebetween by confron-ting bipolar electrodes 130 in electrical series and fluid parallel (i.e. parallel electroly-te feed and parallel electrolyte wi-thdrawal for both anolyte and catholyte)~ -A diffusion barrier separator 145 divides cell 140 into anolyte compartment 141 and ca-tholyte compartment 142.
Similarly, a similar separator 155 divides cell 150 in-to anolyte compartment 151 and catholyte compar-tment 152 and similar separators 165 are provided for the intermediate cells divid-ing them into anoly-te compartmen-ts 161 and catholyte compart-ments 162 While only a small number of cells are shown in the stack for purposes of illustration, i-t will be appreciated that ~: many more cells can be incorporated in such stack on the same principles The end anode electrode 110 comprises a layer 111 of metal which is bonded to a laminated terminal electrode struc- ~
ture comprising conductive ~plas-tic/carbon composite) substra-tes :
- 21 ~
~D'7~35(~1 113, 114 which are sandwiched abou-t an expanded me-tal ~creen Cathode elec-trode 120 comprises conductive sheet 122 with a charcoal coat mounted on a laminated -terminal electrode struc-ture which comprises conductive substrates 123, 124 sandwiched about an expanded me-tal screen 125 Bipolar mid-elec-trodes 130 have metal laye:rs and charcoal layers 132 sandwiched abou-t a central conduc-tive substrate 135 Frame gaskets 143, 144, 153, 15~, 163, 16~ space aid support -the elec-trodes and membranes and comple~te -the enclosure of slab-like electrolyte compar-tments 141, 142, 151, 152, 161, 162, respec-tively, serving as edge walls therefor. The gaske-ts can also provide conduits for feed and return of electroly-te shown in FIG. 2B where gaske-t 143, serving as edge wall for compartment 141 carries feed passage P3 which divides into branch channels 101, 102 to feed anoly-te to upper corners 103, ~ -104 of compartment 141 and re-turn passage P5 which receives anolyte withdrawn from the lower corners of compartment 141 via a similar branched channel arrangemen-t.
The conductive layers 113, 114, 123, 124, 135, and the substrates of cathodes 122, 132 comprise hot pressed mixtures of plastic and carbon, the plastic being polyvinyli-dene fluoride (PVF), polyvinyl chloride, polyethylene or Teflon*, PVF being preferred and the carbon being a finely divided graphite or carbon black The cathode layers 122, 132 comprise charcoa:L pressed into the surface of a carbon plastic substrate which ls in turn bonded to layers 120, 130. The metal layers Illl 131 are hot pressed onto 110, 120, 130 The laminates 113, l:L4, 115 and 123, 124, 125 are separately hot . -~ 30 pressed before adding layers 111 or 112 thereto ~ ' * Registered Trademark : ~ .
~ 935C~
The separa-tors 145, 1~5, :L6~ may be one mil -thick 50~0 porous, elec-trochemical grade polypropylene ~ilm, such as Celgard or 10 mil thick 30% porous Daramic* polyethylene electro-chemical grade ~ilms or other known ion-permeable diffusion-limiting barrier membranes, The end plates 91, 92 and spacer pla-te 93 can be machined or molded to make fluid passages and tie rod holes therein and the marginal edges of the elec-trode, gasket and membrane parts have aligned holes therein. They require appli-cation of sealants prior to assembly, The par-ts are -then assembled as indicated in FIG. 1 and held in compression by tie rods 99 to form a block-like compac-t package, FIGS. 2A-2D are plan views of componen-ts 113, 143, 130 and 93 viewed as shown a-t AA, BB, CC and DD respectively, in FIG, 2, They show the aligned por-ts Pl-P8 running through the marginal portions of those s-tacked components to form longitudinal fluid flow passages. Certain of ports Pl-P8 in end ports are blocked to define longitudinal dead ends of some passages, FIG. 2A shows that sheets 114 and 113 are made of graded resis-tivity so tha-t there is a high resis-tivity between the ports and low resistivity in -the central portion of such sheets where electrode surfaces are bonded, It is important to assure uniformity of elec-troly-te flow through the anolyte and ca-tholyte compartments over the electrode surfaces and to limit short circuit losses between -~
cells, The cell construc-tion minimizes short circui-t losses and enhances uniformi-ty of flow, As shown in FIGS. 2, 2B
~ : ,. .
and 2C wherein a dis-tribution channel 141DC is provided in anolyte compartment ~41 and a distribu-tion channel 142FM is provided in the catholyte compartment 142. These channels * Registered Trademark ':
~'7~350 serve as plenum chambers and are formed in-tegrally wi-th -the compar-tmen-t struc-ture. Similar return channels are provided for the respec-tive compartments in cells 140, 150 and 160, FIG, 2B indica-tes flow pat-tern (arrows F) for anolyte compart-ment 141, The feed passage in P3 branches into lateral delivery channels 101, 102 with respective corner exits 103, 104 at ends of channel 141DC at -the upper corners of compartment 141, Typical cross-sec-tion dimensions for channel 141DC or 142DC are 1/16 X 1/8 inch (or more in the case of 141DC de-pending on the thickness of metal layer employed) and ~the thick-ness of the central portion of compar-tmen-ts 141 and/or 142, established by separation of the anode or ca-thode surfaces from -the confron-ting surface of membrane 145, is typically 1/64 inch. The distribution channel and plenum chamber flow pattern is thus intrinsically established with minimal ex-traneous structure for this purpose and consis-tent wi-th multiple spaced point fluid entry (or drain) at the cell compar-tment edge which allows a significant design length of parallel electrolyte flow paths (the length counterbalancing low electrolyte re-sistivity to establish an acceptably high in-tercell resistance) t~ be establiShed in a small space within the overlapping margin of component cells of the cell stack.
Spacers (indicated for example at 158, 159) are set in the anolyte and ca-tholyte compartments of each cell to es-tablish a uniform spacing of anode and cathode electrodes from the thin barriers 145, 155, 165. The spacers comprise netting with raised crossovers between intersecting synthetic fiber threads (e.g. polypropylene), Alternatively, the net can be pressed into a corrugated form to permit fluid flo~ while maintaining said spacing.
Referring to FI~. 2D, transverse channels D15, D26, D37 and D48 are formed as troughs in plate 93 and they butt . . , ~ .. ~.. .
3~
against subs-tra-te 124 -to ~orm comple-te elongated passages Pl to P5, P2 -to P6, P3 -to P7 and P4 to P8. Passages Pl, P2, P7 and P8 may con-tinue -through pla-te 93 -to ~eed a further array of cells SB2 utilizing common anoly-te and catholy-te recirculating means for both SBl and 2.
- FIG. 2B is a plan view of the spacing gaske-t frame 143 It is made of insulating material, e.g. polyvinyl chloride and is provided with ports Pl-P8 which are punched through during or after frame manufac-ture. If ports Pl-P8 are formed af-ter frame manufac-ture, only a single molding die is necessary to form -the shapes of all frames 143, 144, 163, 164, 123, 124 and the -troughs 101, 102 (upper and lower) therein and a single multiple hole punch can form ports Pl-P8 for all overlapping parts 113, 114, 123, 124, 143, 144, 145, 153, 154, 155, 163, 164, 165. The metal screens 115 and 125 would be short enough to be by-passed by ports Pl-P8 and, in practice, would extend ou-t of electrodes 110 and 120 in a direction ro-tated 90 from the direction shown in FIG 2 for convenience of illustration. -FIG. 2B shows anolyte feed through P3 and drain through P5 and catholyte feed would be in P4 and drain in P6 However, it is useful in some instances to have cross-over, i.e.- anolyte feed at P3, drain at P6 and ca-tholy-te feed at P4, drain at P5, in order -to balance -their flow pa-tterns in the cell compartments by compensating flow resistance in their respective tr1butary feed and drain passages FIG 2C indicates a plan view of the bipolar electrode laminate 130. The raised anode surface 131 is so proportioned in size that its side edges can be readily conformed to sides of frame 144 to prevent side leakage of electrolyte around the electrode.
FIGS 2, 2B and 2C show that the raised elec-trodes, ~ .
~ ~ - . . .
9~5~
e,g., anode layer 131 in bipolar laminate 130 have ~ lesser heigh-t -than -the compar-tmen-ts and are positioned -to form the mani~olds (e.g,, 141DC, 142DC of FIG, 2) .
Typical manufacturing condi~tions which have ~een used ~or making cells of the type shown in FIG, 1 are:
(1,0) for making sheets 114, 124, (1,1) Mix 50~o 3548 Kynar~ (P~F copolymer) wi-th 50%
Dixon Grade 1112 graphite, Sieve PVF copolymer to break up clumps, (1.2) Blend for 3 minu-tes in PK blender wi-thout agi-tator bar plus 3 minutes with agi-ta-tor bar.
(1,3) Sift to break up clumps.
(1,4) Trowel ou-t a slab of this mix-ture 0,100 in.
-thick, However, apply a mix-ture of 60% PVF, 40~o graphite in -the end regions around por-ts Pl-P8 -to achieve resis-tivity graduation as described above in connection with FIG, 2~, (1,5) Preheat for 7 minutes at 350F, (1,6) Press for 7 minutes at 400F at 256 psi, (1.7) Cool for 7 minutes at 50F a-t 256 psi, (1, 8) Release (2,0) ~or making sheets 113, 135 and 123~ -(2,1) Mix 40% PVF copolymer Kynar* grade 3584 with 60~o Dixon Grade 1112 graphite. Sieve copolymer to break up ;; clumps, (2,2) Blend as in (1,2), ; ~ (2,3) Si~t as in (1,3), (2,4) Trowel a slab as in (1,4), (2,5) Preheat as in (1, 5), (2,6 3 Press as in (1,6).
., (2,7) Oool as in (1.7).
(2,8) Release * Registered Trademark ::
~ ~ _ 26 -- ~ .. : :, , , (3.0) for making ca-thodes 132 and 122, (3,1) ~ix:
55% Dixon Grade 1175 graphi-te 5% Dixon Grade XC 72 graphite which has been pulverized to go through a 40 mesh screen 40% PVF copolymer (Kynar* grade 3584) (3,2) Blend (PK blender withou-t agitator bar) 6 minutes (3.3) Sif-t to break up clumps (3.4) Trowel a slab 3/16 in. thick (3,5) Prehea-t at 400F for 7 minu-tes (3, 6) Press a-t 400F, 256 psi for 7 minu-tes (3,7) Cool a-t 50F~ 256 psi for 7 minutes (3,8) Release (3.10) To coat sheets 123, 135 with surface area enhancing layer, (coatings 122, 132) (3,11) Using activated charcoal such as Darco* 20/40, spread a coating 1/8 inch thick on the surface of the sheet (3,12) Preheat the coating and the sheet -to 400F
20 ~or 3 minutes (3.13) Press at 400F for 3 minu-tes at pressure of 200 psi (3,14) Cool at 50 for 3 minutes at a pressure o~
200 psi (3,15) Release.
.
(4,0) To laminate sheets 114 and 113 with copper conductor 115 sheets 123 and I24 with copper conductor 125, (4,1) Sprinkle half complete mo~olayer of mixture of ~; ~ 60~o grade 1112 g:raphite between the copper screen and the 3a ~ two sheets, :~ :
~ ~ * Registered Trademarks .
35~
(4,2) Preheat at 400F for 3 minutes, (4,3) Press a-t 200 psi a-t L~oo F :~or 3 minutes, (4,4) Cool at 50F at 200 psi fo:r 3 minu-tes, The di~fering graphite loading layers 113 and 114 (and of 123 and 124) throughout the major central portions -thereof balance coefficients o~ thermal expansion and produce a flat resultant laminate which is important ~or later assembly and reliable cell dimensioning, (5,0) To make anode sur~aces 111, :L31 sui-table ~or ion plating, (5,1) Trowel a 1/8 inch -thick layer o~ 50 to 200 mesh coconut charcoal onto a conductive shee-t made as in 2,0 or 3,0 above, (5,2) Preheat a-t 400F for 3 minu-tes, (5,3) Press a-t 400F and 200 psi for 3 minu-tes, (5,4) Cool at 50F and 200 psi for 3 minutes, (6,o) To laminate the coated cathode elec-trodes made per (3,0) above, the anode electrodes made per (3,0) above, the assemblies made per (4,0) above and -to the sheets 130 (2,0), (6,1) Assemble layers and prehea-t the assembled parts 350F for 3 minutes, (6,2) Press at 350F for 3 minutes a-t 80 psi, (6,3) Cool under pressure and release~
(6,10) The passages Pl to P8 and tributaries which carry con~ ~
ductive electrolyte need to be insulated as they pass -through ~ -the elec-trodes, This is accomplished by resistivity grading as described above or as follows, (6,11) Mix 95% PVF copolymer (Kynar* 3584~ wi-th 5%
graphite (Dixon Grade 1112), (6,12) Trowel a slab 3/16 -thick, (6,13) Preheat at 400F for 7 minu-tes, (6,14j Press at 400~ for 7 minutes at 135 psi, -Registered Trademark ' 3~
(6.1~) Cool a-t 50F f:or 7 minu-tes at 135 psi, (6,16) Cut into plugs B 1/6~ smaller in dlameter -than the holes in the elec-trodes -themselves, (6.17) Insert plugs B (FIG. 2C) in-to -the holes in the elec-trode shee-t.
(6.18) Preheat electrode sheet and the plug to ~00~
for 7 minutes.
(6,19) Press at 400 F for 7 minu-tes a-t 116 psi.
(6,20) Cool a-t 50F for 7 minutes at 116 psi.
(6.21) Release, This -technique may also be used for elec-trodes 110 and 120 in lieu of resis-tivity grading as described above, (7,0) Fluid passages 101, 102, e-tc,, are cut as grooves into gaske-ts 1~3, 14L~, 153, 154, 163, 164, (8,0) S-tacks of cells are made up and in so doing the mating surfaces are coa-ted with a sealant/adhesive (with coa-tings applied on grooved gasket faces above and below the groove --e,g. above and below groove 101/102 in FIG. 2B) to prevent liquid leakage, The grooved gasket faces butt against the mem-branes 145, 155, 165. Pressure plates (91, 92, 93) are applied at the ends of the stack and tied together with long bolts 99 to compressively grip the cell stack together in a manner well known in the art.
The anolyte and catholyte supplies (anolyte in and catholyte in) are indicated at A,I, and C.I., respectively and the (spent) anolyte and catholy-te drains a-t A,0. and C,0.
respectively (anolyte out, catholyte out).
Therefore, anolyte comes into the stack through longi-tudinal channel P7, and catholyte comes in through longitudinal channel P8 of p3.ate 91. The ~nolyte and catholyte continue in these ~channe]s through SBl to press plate 93 where the . .
anolyte traverses transverse channel 37 to P3 and the catholyte , .
3~
traverses -trans~erse channel D~8 -to P4 The anolyte and ca-tholy-te -travel back longitudinally -through SBl via channels P3 and pl~ respec-tively where -they are dis-tributed to the uni-t cells via ~branch tribu-taries as des-cribed in connec-tion wi-th FIGS. 2B and 2 above. Af-ter passing through the unit cells, and genera-ting elec-trical power -there, the anolyte and catholyte emerge, via branched -tributaries, to respective return pa-ths P~ and P6, then -traverse transverse channels 26 and 15, respectively, then respectively traverse longitudinal paths Pl and P2, an.d then re-emerge from the cell stack on -the ou-tside of plate 91, as indica-ted by -the lines A 0. and CØ
Separator screens (not shown) are included in each of the anolyte and catholyte compar~tmen-ts to preven-t the flexible diffusion barrier from deflecting over into contact with other elec-trodes confronting it Suitable screens include Dupont's Vexar* Model 10 PDS 169, which has high density poly-ethylene strands of 10 mil width criss-crossing in a diamond pattern with 16 strands per inch and affording 90~0 porosity and which can be corrugated -to various dep-ths of corrugation for fitting in different volumes of Hercules' Delnet* brand embossed and stretched polypropylene sheet, Model GQ 330, affording a 75% porosity and having a thickness o~ 10-12 mils and raised but-ton embossed projections to any desired separation :.
depth The anolyte and catholyte compartments may be di~er-entially pressurized to establish a bias agains-t diffusion of undesired ionic species. Such differential pressurization can :
be achieved through different respective pump sizes and/or :
~ -30 speeds, flow passage sizes; or through throttle valves. ~ :
: ' * Registered Trademark .
11 ~37935i~
An alterna-tive or supplemen-t -to pressure differential diffusion suppression is -the use of a mi~-compartment (not shown) in each uni-t fuel cell between anolyte and catholy-te compartments and separated -therefrom by diffusion barriers or coarser colloid barriers. Such mid-compartments, if used, may be provided with their own recirculating sys-tems with filters -to trap certain diffusing species or may be pressurized to suppress in ei-ther direction.
Internal resistances in the converter may comprise the following components:
resistivity of anolyte: e.g. 2-6 ohm-in. for initial 20%
NaCl in water average over operation of a converter resistivity of catholyte: e.g. 4-10 ohm-in. for 4 Molar FeC13 in water average over operation of a converter resistance of 60 sq in. separator: about.O5 ohm-in. for .010"
thick 30% DARAMIC saturated with 25% NaCl solution or equivalent Other losses are pumping requirements to move electrolytes. In a typical case for 60 sq. in. electrodes at 0.5 amp/sq. in. current drain the pumps move anolyte at O -to 1 cc/sec./cell, catholyte at 1 to 2 cc/sec./cell under a 8 to 10 foot head with a pressure differential between anolyte and -catholyte corresponding to 1-2 feet of head.
~ Spacing in the electrolyte compartments for purposes ; of resistance calculation chanees during a discharge. Anode to barrier spacing grows from .015 inches as metal deposited earlier during charge is consumed FLO~ RATE
2 cc/sec. of 4M FeC13 corresponds to a reagent supply rate of .008 equivalent weights per second or about 30 ew/hour.
35~
At 30A (1/2A/in.2) one consumes jus-t over 1 ew/hr. (27 AH = lew).
So supply ra-te is -then abou-t 30 times -the ~,toichrometric requirement and, in any even-t, should be at least lOX. Since Fe +3 concen-tration, goes ~rom 4M to abou-t .5M There is a one--tenth of a molarity charge per pass.
Example 1 A 60 sq. in nominal electrode area single cell was made by assembling a flat sheet o~ armco iron, a terminal cathode prepared as in (6.o) as above (but using Calgon Corpor-ation BP~ grade charcoal) and a .010 inch thick Daramic poly-ethylene separator with 075 inch thick rectangular frame gaskets provided with in-tegral electrolyte feed channels. An anolyte comprising 125 ml. of 20% (by weight) NH4Cl in water was circulated through the anolyte compartmen-t a-t a rate of 2 cc/sec. and a catholyte comprising 500 ml. of 4 Molar FeC13 in water was circulated through the catholyte compar-tment at a rate of 2 cc/sec.
The cell exhibi-ted an open circuit potential of 1 100 v. and gave the following performance with time at 15 7 ampere drain:
Min V V
Time 1 d . 1.10 ,, ' 1 .73 1 03 .69 .94 .65 .87 .60 80 .53 .76 .13 32 81 minutes .62 85 As used above and hereinafter ~1 is load voltage and Vd is driving potential, the latter being measurable by momentarily : ~ , - .
.
:'
35~
At 30A (1/2A/in.2) one consumes jus-t over 1 ew/hr. (27 AH = lew).
So supply ra-te is -then abou-t 30 times -the ~,toichrometric requirement and, in any even-t, should be at least lOX. Since Fe +3 concen-tration, goes ~rom 4M to abou-t .5M There is a one--tenth of a molarity charge per pass.
Example 1 A 60 sq. in nominal electrode area single cell was made by assembling a flat sheet o~ armco iron, a terminal cathode prepared as in (6.o) as above (but using Calgon Corpor-ation BP~ grade charcoal) and a .010 inch thick Daramic poly-ethylene separator with 075 inch thick rectangular frame gaskets provided with in-tegral electrolyte feed channels. An anolyte comprising 125 ml. of 20% (by weight) NH4Cl in water was circulated through the anolyte compartmen-t a-t a rate of 2 cc/sec. and a catholyte comprising 500 ml. of 4 Molar FeC13 in water was circulated through the catholyte compar-tment at a rate of 2 cc/sec.
The cell exhibi-ted an open circuit potential of 1 100 v. and gave the following performance with time at 15 7 ampere drain:
Min V V
Time 1 d . 1.10 ,, ' 1 .73 1 03 .69 .94 .65 .87 .60 80 .53 .76 .13 32 81 minutes .62 85 As used above and hereinafter ~1 is load voltage and Vd is driving potential, the latter being measurable by momentarily : ~ , - .
.
:'
3 5 opening the load circui-t.
Example 2 Another cell was prepared as in Example 1 with the following changes:
Cathode - 40 to 80 mesh 0~ charcoa] on a carbon/plas-tic com-posite substra-te. Vexar screen crimped -to .065 on anode side, all other constructional features being the same. The cell was charged at 15A with 3.5 Molar ~eC13 circulated through anolyte compartment (500 ml). ~he charging vol-tage was l,L~8 -1.52 over 66 min at 15 amperes. Ins-tan-taneous open circuit vol-tage of 1.22 - 1.26 (Back E~F) was measured from time while charging. After charging the system was idle at 1.16 VOC with the same electrolytes. The system was discharged at 16,4 amps, VD was 1.06 to 1.02 and VL was 80 to .76 over 5 min. Then fresh NaCl anolyte was added and VD was 1.06 to 0.96 and VL was .81 to ,72 over 4 minutes. Then fresh 4M FeC13 was added and VD was 1.08 at 16,4 amps and a flat V~ was .84; Vo was 1.04 at 20 amps; V~ was .79 flat; and VD was .99 to .30A; V~ was .63, Example A 60 in2 cell 2 with energy average of 2AH/in was made as in Example 2 excepting the Anode had some iron in its surface before charging and -that .055 in frames and crimped screens were used. ~he cell for 67 minutes at 15A amperes at 1,5 volts produclng 1.21 Voc, discharged for 5 minutes at 16,6 ;~
amperes, .86 vol-ts (10% NaCl, 2M FeC12); charged for 72 minutes at 15 amperes, 1,52 volts (1.18)(34 AH total); discharged for 5 minutes at 16,2A, .80V. (1.08) using electrolytes from dis-charge above: charged for 70 minutes at 15A, 1,55V. (1.22) using electrolytes from charge above (51AH); discharge for 6 ;~
3o mlnutes at 17A, .87V. (1.16) with fresh solutions; charged for 56 minutes a-t 15A, 153V, (64AH total); discharge for 3 minutes at 16,8A, .88V. (1,74) with electrolytes used above.
~: :
, .
- 33 - ~
' ....
1S~ 3~
Charged for 52 minutes a-t 15A, 1.6V. (1.25) 76A~I total;
discharge for 10 minu-tes at 16.6A, .8L~ -to 80 (l 14 - 1.10) same elec-troly-tes as before; charge 87 minu-tes at l~A, 1.6V.
(1.27) 96 AH -to-tal; discharge for ~ minutes at 16.5 0.84 (1 10);
charge for 45 minutes at 15A, 1.59V. (1.25); idle 10 minutes Voc was 1.12; discharge 18 minutes a-t 16A, .84 (1.08) to .77 (1.08) old electrolytes; fresh NaCl, (1.12); idle to 1.20;
fresh FeC13 .83 (1 18); charge 96 minu-tes at 15A, 1.65 (1 3L~);
fresh elec-trolytes 1 46 (1.02); total of 120 AH in.
In summary, the cell was charged -to a total o~ 2A~I/in2 at a rate of 1/4 A/in2; charging vol~tages ranged from 1.50 to 1 65V (open circui-t vol-tages from 1.18 to 1.3L~V); 5 to 10 minu-te discharges during -the charging process gave constan-t performance -through all sta-tes of charge indicating -that there was no degradation of ~the surface as capacity was increased and that the system is highly -tolerant of cycle ~ariations.
Example 4 Cathode: 20 to 40 OL charcoal with no substrate Anode: 50 to 200 coconut charcoal wi-th no substrate Vexar screens crimped to .o65 in bo-th compar-tments;
10 inch Daramic barrier; Anolyte: 300 ml of 4M FeC12;
Catholyte: 500 ml of 4M FeC12; charged for 132 minutes at 15A
or 33 A~; charging voltage 1.52 -to 1.62V. Both EMF 1.21 to 1.31 Yolts idle to 1.16.
Fresh anoly-te (10% ~aCl) and fresh catholyte 4M FeC13 were added -to produce an open circuit vol-tage of 1.23 volts.
~he cell was discharged at 16.4A, .84 volts (1.18 VD flat) for 12 minutes, -then at 30A, 60Vl, 1.03 - 1 13VD (i.e., .014 to 017R) Example ~:
Anode Polarization C ar~in~
':
,, , , ' ~ : .-: . . . .
3~
4M FeC12 anolyte (.5e) and ca-tho:Ly-te (le,); charged at 20A ~or 96 minu-tes (32AH); charging voltage 1,58 - 1,60V;
open circui-t (instantaneous) 1,20 -t;o 1,23, DischarFin~
A, (2M ~eC13 catholyte) Anol,yte Voc VL VD VL VD
. ~
1,3 M FeC12 ,17 ,89 (1,16) ,88 (1,15) 16,7A
2 M FeC121,15 ,85 (1,12) ,84 (1,10) 16,6A
2.7 M 1,11 ,79 (1.08) ,77 (1.07) 16,7A
B, 2M FeC13 catholyte ~00 ml 1,3 M FeC12 1,17 ,89 (1,16) ,88 (1,15) 2 M FeC12 1,15 ,85 (1,12) ,84 (1,10) 2,7 M FeC12 1,11 ,79 (1,08) ,77 (1,07) Fresh cathol~te at room temperature 1,3 M FeC12 1,21 ,78 (1,14) 6 M FeC12 1,16(1) ,75 (1,11) (3)-,50 (r95j ~ ;
2 M FeC12 1,15 ,7O (1,14) 6 M FeC12 1,14(1) ,72 (1,10) (4)-,50 (1,00) Performance is only slightly impaired by ferrous ion concentra-tions up to 4M; however, 6M ferrous chloride anolyte causes severe polarization of the anode, Since the anolyte is near saturated, additional ferrous chloride produced at the anode apparently crystallizes out on the surface, FIG, 3 shows another embodiment of the invention com-prising a multi-cell energy conversion array package 310, two ; magnetically coupled centrifugal pumps Pa and Pb for anolyte and catholyte, two flow rate control valves, Vl and V2, in ~ ~ series with the pump, two solid particle line fil-ters, Fa and ; ~ Fb, four - two way valves, A, B,~ C and D for switching filters ~ -; 30 from one hydraulic circuit to another, Interconnected anolyte ; and catholyte reservoir tanks AT and CT, two drain cocks, DC, , for flushing the system, Valves A, B, C and D are -two way _ ':, 3~ .
switches which are employed -to transfer filters Fa and Fb from anoly-te -to catholy-te respectively and vice-versa, However, these valves also permi-t -the flow of elec-trolyte from one -tank to another depending upon the combination of positions.
In the normally operatirlg mode the valves are in the following positions.
Valve A Position ~1 or ~2 Valve B position ~1 or #2 Valve C position -~2 or ~1 Valve D posi-tion #2 or ~1 In the firs-t set of positions fil-ter Fa is in the anolyte circuit and Fb is in the ca-tholyte circui-t, With the valves in the second set of positions -the filters are placed in the opposi-te circuits.
When it becomes necessary -to -transfer some anolyte to the catholyte circuit or vice-versa the valves are set in the appropriate positions, For example, with the valves in the following positions;
Valve A position #l Valve ~ position #l ; Valve C~ position #l Valve D position #2 electrolyte flow is such tha-t the catholyte does not drain .
out of its tank while the anoly-te not only flows through Fa into its tank, but also through Fb, -through the catholy-te compartment and into the catholyte tank, There is thus a net ~ transfer of anolyte into the catholyte reservoir via the paths ; described. If valves Vc and Vd are put into positions #2 and #l respectively, catholyte is transferred to -the anolyte tank, .
The slxteen available position combinations provide a large number of operational modes, An available circui-t for ~, mixîng bo-th electrolytes completely is -the valve position ~ ~ .
~7~3~i~
combination, VA posi-tion t~2 VB position ~1 VC position #
VD position ~2 In this configuration, shown in figure 4, all circuits become one large, single series circui-t.
The electrolytes become a single fluid within a fairly short -time.
Both circulating pumps are located below the tes-t s-tand platform to enable them to be primed rapidly. Drain cocks are provided near the exit connection to both tanks so that the system can be easily drained and flushed, or it can be used to obtain periodic samples of elec-troly-te for examina-tion or analysis. -~
Fluid tanks can be interconnected so that any net flow oP electrolyte via diffusion across the porous barrier can be compensated by automatic spill over in-to the appropriate tank. In -this manner -there is no chance of tank overflow or total depletion of anolyte or catholyte during long tes-t periods. The tank volumes are designed for the maximum quantities of fluid one may practically encounter in the operation of a 10 cell array.
The anolyte tank has about a 2 to 3 liter usable volume and the catholyte tank about 6 gallons or 24 liters. ~
For single cell testing separate 2 to 4 liter tanks may be ~ ;
,.
connected to the test stand, or, it is possible to employ the fabrica-ted tanks with very low liquid levels. The tanks are vented to the air ~and at the immediate exit por-t of -the cells to vent gases and to minimize the chance of creating undue ~ pressure differentials across the barrier or with respect to :: :
the cell input and output `~
: '' :
.. ~.. , ~ . .. . ... , : .. .
~7.~3~
Fil-ters are AMF Cuno~ wa-ter ~ ters with cartridges designed specifically to remove par-ticula-te ma-tter o~ -the size and form of iron oxides, These fil-ters are all plastic in cons-truction and easily dismantled ~or car-tridge replacement, They hold about one liter o~ ~luid and represent a sizeable portion of -the total fluid for small devices, These filters can be reduced in size by cutting off the lower por-tion of the housing to a desired length, cementing a new base and -then trDmming filter cartridges to fit the shor-ter length, FIG, 5 shows a further embodiment of -the invention in which a ~uel cell 500 comprises a-t least a one-ce]l (pre-ferably more) cons-truc-tion in a major energy conversion device with an anolyte compar-tmen-t 501 and ca-tholyte compar-tment 502, both of which are fed by pumps 503, The anolyte flows in a recirculating closed loop comprising s-torage tank 504 and filter Fa, The catholyte flows in a recirculating closed loop which includes filter F6 and a reservoir 505, An auxiliary energy conversion device 510 comprises an anolyte compartment 511 and a catholy-te compartment 512 receiving bIeed flows of less than lO~o prefereably less than 5~0 by volume throughou-t, from the recirculating main electrolyte flows, The anolyte of the main cell is catholyte in the aux-iliary cell and the catholyte of -the main cell is anolyte in the auxiliary cell. The auxiliary cell is electrically powered, preferably by trapping the power generation of the main cell to es-tablish an overpotential across the auxiliary cell for overcharging to chlorine potential (2,o volts or higher), Example 6 ;::
Charge-discharge data was plotted as shown in FIG, 6 * Registered Trademark "''~ ' .
., - . .
s~
for a lL~ cell array having :L50 -to 200 m:il thick frames which had feeder holes drilled into edges of -the plates to establish a leak free design. The array was assembled via gaskets and bolted and plates. The anoly-te volume was abou-t hal~ a gallon, catholyte volume about 1.3 gallons and -the molari-ty of the FeC12 solution was 3 to 3 molar. The theoretical maximum capacity of unit as de-termined by 1 3 gal. of 3 2 molar is 1.3 X 4 liters = 5.2 liters of 3.2 molar. This is equivalent to 5 2 X 3.2 = 16.6 moles of FeC12. A-t a maximum o~ 17 ampere-hours per mole, one ob-tains 17 X 16 6 = 283 AH total for the array. Since -there are 14 cells in -the array, each cell could receive 283/14 = 20.2 AH to-tal, if the reagent utiliza-tion factors were lOO~o. No fil-ters or pH adjus-tment apparatus were employed Charging the array a-t 15 amps for 2 hours gave an input of 30 AH. The output was 20 AH at 10 amp rate. Com-plete utilization of electrolyte was realized. Overcharging is evidenced by rise in open circuit voltage and charging voltage towards end of charge mode. Polarization losses are evidenced at this molarity range and dep-th of discharge and curren-t densities by differences in o.c.v. for charge and discharge modes Example 7 The Example 6 14 cell array was operated under same conditions of electrolyte volumes and molarity, etc. as in ~; Example 6, excepting that charging was at 10 amperes and dis-charge at 8 amperes. Results for -two such cycles are shown in FIGS. 7 and 8. The reagent utilization and current efficiency appear quite good, In the FIG, 7 run, there were 5 AH input ..
and almos-t 4 AH output. In the FIG. 8 ru~n, the input was well over 8 AH. Polarizatlon effects, e.g., the differences in - o.c v, of charge and discharge, appear less a-t these lower :
~ 3tj~
curren-t densi-ties compared -to Example 6, Example 8 A 12 cell array assembled in same ~ashion as -the 14 cell array of Examples 6-7, The elec-trolyte was 3 molar FeC12 initially and catholyte volume was about 4 gallons and anolyte 1 gallon. The coulombic capacity of this system is
Example 2 Another cell was prepared as in Example 1 with the following changes:
Cathode - 40 to 80 mesh 0~ charcoa] on a carbon/plas-tic com-posite substra-te. Vexar screen crimped -to .065 on anode side, all other constructional features being the same. The cell was charged at 15A with 3.5 Molar ~eC13 circulated through anolyte compartment (500 ml). ~he charging vol-tage was l,L~8 -1.52 over 66 min at 15 amperes. Ins-tan-taneous open circuit vol-tage of 1.22 - 1.26 (Back E~F) was measured from time while charging. After charging the system was idle at 1.16 VOC with the same electrolytes. The system was discharged at 16,4 amps, VD was 1.06 to 1.02 and VL was 80 to .76 over 5 min. Then fresh NaCl anolyte was added and VD was 1.06 to 0.96 and VL was .81 to ,72 over 4 minutes. Then fresh 4M FeC13 was added and VD was 1.08 at 16,4 amps and a flat V~ was .84; Vo was 1.04 at 20 amps; V~ was .79 flat; and VD was .99 to .30A; V~ was .63, Example A 60 in2 cell 2 with energy average of 2AH/in was made as in Example 2 excepting the Anode had some iron in its surface before charging and -that .055 in frames and crimped screens were used. ~he cell for 67 minutes at 15A amperes at 1,5 volts produclng 1.21 Voc, discharged for 5 minutes at 16,6 ;~
amperes, .86 vol-ts (10% NaCl, 2M FeC12); charged for 72 minutes at 15 amperes, 1,52 volts (1.18)(34 AH total); discharged for 5 minutes at 16,2A, .80V. (1.08) using electrolytes from dis-charge above: charged for 70 minutes at 15A, 1,55V. (1.22) using electrolytes from charge above (51AH); discharge for 6 ;~
3o mlnutes at 17A, .87V. (1.16) with fresh solutions; charged for 56 minutes a-t 15A, 153V, (64AH total); discharge for 3 minutes at 16,8A, .88V. (1,74) with electrolytes used above.
~: :
, .
- 33 - ~
' ....
1S~ 3~
Charged for 52 minutes a-t 15A, 1.6V. (1.25) 76A~I total;
discharge for 10 minu-tes at 16.6A, .8L~ -to 80 (l 14 - 1.10) same elec-troly-tes as before; charge 87 minu-tes at l~A, 1.6V.
(1.27) 96 AH -to-tal; discharge for ~ minutes at 16.5 0.84 (1 10);
charge for 45 minutes at 15A, 1.59V. (1.25); idle 10 minutes Voc was 1.12; discharge 18 minutes a-t 16A, .84 (1.08) to .77 (1.08) old electrolytes; fresh NaCl, (1.12); idle to 1.20;
fresh FeC13 .83 (1 18); charge 96 minu-tes at 15A, 1.65 (1 3L~);
fresh elec-trolytes 1 46 (1.02); total of 120 AH in.
In summary, the cell was charged -to a total o~ 2A~I/in2 at a rate of 1/4 A/in2; charging vol~tages ranged from 1.50 to 1 65V (open circui-t vol-tages from 1.18 to 1.3L~V); 5 to 10 minu-te discharges during -the charging process gave constan-t performance -through all sta-tes of charge indicating -that there was no degradation of ~the surface as capacity was increased and that the system is highly -tolerant of cycle ~ariations.
Example 4 Cathode: 20 to 40 OL charcoal with no substrate Anode: 50 to 200 coconut charcoal wi-th no substrate Vexar screens crimped to .o65 in bo-th compar-tments;
10 inch Daramic barrier; Anolyte: 300 ml of 4M FeC12;
Catholyte: 500 ml of 4M FeC12; charged for 132 minutes at 15A
or 33 A~; charging voltage 1.52 -to 1.62V. Both EMF 1.21 to 1.31 Yolts idle to 1.16.
Fresh anoly-te (10% ~aCl) and fresh catholyte 4M FeC13 were added -to produce an open circuit vol-tage of 1.23 volts.
~he cell was discharged at 16.4A, .84 volts (1.18 VD flat) for 12 minutes, -then at 30A, 60Vl, 1.03 - 1 13VD (i.e., .014 to 017R) Example ~:
Anode Polarization C ar~in~
':
,, , , ' ~ : .-: . . . .
3~
4M FeC12 anolyte (.5e) and ca-tho:Ly-te (le,); charged at 20A ~or 96 minu-tes (32AH); charging voltage 1,58 - 1,60V;
open circui-t (instantaneous) 1,20 -t;o 1,23, DischarFin~
A, (2M ~eC13 catholyte) Anol,yte Voc VL VD VL VD
. ~
1,3 M FeC12 ,17 ,89 (1,16) ,88 (1,15) 16,7A
2 M FeC121,15 ,85 (1,12) ,84 (1,10) 16,6A
2.7 M 1,11 ,79 (1.08) ,77 (1.07) 16,7A
B, 2M FeC13 catholyte ~00 ml 1,3 M FeC12 1,17 ,89 (1,16) ,88 (1,15) 2 M FeC12 1,15 ,85 (1,12) ,84 (1,10) 2,7 M FeC12 1,11 ,79 (1,08) ,77 (1,07) Fresh cathol~te at room temperature 1,3 M FeC12 1,21 ,78 (1,14) 6 M FeC12 1,16(1) ,75 (1,11) (3)-,50 (r95j ~ ;
2 M FeC12 1,15 ,7O (1,14) 6 M FeC12 1,14(1) ,72 (1,10) (4)-,50 (1,00) Performance is only slightly impaired by ferrous ion concentra-tions up to 4M; however, 6M ferrous chloride anolyte causes severe polarization of the anode, Since the anolyte is near saturated, additional ferrous chloride produced at the anode apparently crystallizes out on the surface, FIG, 3 shows another embodiment of the invention com-prising a multi-cell energy conversion array package 310, two ; magnetically coupled centrifugal pumps Pa and Pb for anolyte and catholyte, two flow rate control valves, Vl and V2, in ~ ~ series with the pump, two solid particle line fil-ters, Fa and ; ~ Fb, four - two way valves, A, B,~ C and D for switching filters ~ -; 30 from one hydraulic circuit to another, Interconnected anolyte ; and catholyte reservoir tanks AT and CT, two drain cocks, DC, , for flushing the system, Valves A, B, C and D are -two way _ ':, 3~ .
switches which are employed -to transfer filters Fa and Fb from anoly-te -to catholy-te respectively and vice-versa, However, these valves also permi-t -the flow of elec-trolyte from one -tank to another depending upon the combination of positions.
In the normally operatirlg mode the valves are in the following positions.
Valve A Position ~1 or ~2 Valve B position ~1 or #2 Valve C position -~2 or ~1 Valve D posi-tion #2 or ~1 In the firs-t set of positions fil-ter Fa is in the anolyte circuit and Fb is in the ca-tholyte circui-t, With the valves in the second set of positions -the filters are placed in the opposi-te circuits.
When it becomes necessary -to -transfer some anolyte to the catholyte circuit or vice-versa the valves are set in the appropriate positions, For example, with the valves in the following positions;
Valve A position #l Valve ~ position #l ; Valve C~ position #l Valve D position #2 electrolyte flow is such tha-t the catholyte does not drain .
out of its tank while the anoly-te not only flows through Fa into its tank, but also through Fb, -through the catholy-te compartment and into the catholyte tank, There is thus a net ~ transfer of anolyte into the catholyte reservoir via the paths ; described. If valves Vc and Vd are put into positions #2 and #l respectively, catholyte is transferred to -the anolyte tank, .
The slxteen available position combinations provide a large number of operational modes, An available circui-t for ~, mixîng bo-th electrolytes completely is -the valve position ~ ~ .
~7~3~i~
combination, VA posi-tion t~2 VB position ~1 VC position #
VD position ~2 In this configuration, shown in figure 4, all circuits become one large, single series circui-t.
The electrolytes become a single fluid within a fairly short -time.
Both circulating pumps are located below the tes-t s-tand platform to enable them to be primed rapidly. Drain cocks are provided near the exit connection to both tanks so that the system can be easily drained and flushed, or it can be used to obtain periodic samples of elec-troly-te for examina-tion or analysis. -~
Fluid tanks can be interconnected so that any net flow oP electrolyte via diffusion across the porous barrier can be compensated by automatic spill over in-to the appropriate tank. In -this manner -there is no chance of tank overflow or total depletion of anolyte or catholyte during long tes-t periods. The tank volumes are designed for the maximum quantities of fluid one may practically encounter in the operation of a 10 cell array.
The anolyte tank has about a 2 to 3 liter usable volume and the catholyte tank about 6 gallons or 24 liters. ~
For single cell testing separate 2 to 4 liter tanks may be ~ ;
,.
connected to the test stand, or, it is possible to employ the fabrica-ted tanks with very low liquid levels. The tanks are vented to the air ~and at the immediate exit por-t of -the cells to vent gases and to minimize the chance of creating undue ~ pressure differentials across the barrier or with respect to :: :
the cell input and output `~
: '' :
.. ~.. , ~ . .. . ... , : .. .
~7.~3~
Fil-ters are AMF Cuno~ wa-ter ~ ters with cartridges designed specifically to remove par-ticula-te ma-tter o~ -the size and form of iron oxides, These fil-ters are all plastic in cons-truction and easily dismantled ~or car-tridge replacement, They hold about one liter o~ ~luid and represent a sizeable portion of -the total fluid for small devices, These filters can be reduced in size by cutting off the lower por-tion of the housing to a desired length, cementing a new base and -then trDmming filter cartridges to fit the shor-ter length, FIG, 5 shows a further embodiment of -the invention in which a ~uel cell 500 comprises a-t least a one-ce]l (pre-ferably more) cons-truc-tion in a major energy conversion device with an anolyte compar-tmen-t 501 and ca-tholyte compar-tment 502, both of which are fed by pumps 503, The anolyte flows in a recirculating closed loop comprising s-torage tank 504 and filter Fa, The catholyte flows in a recirculating closed loop which includes filter F6 and a reservoir 505, An auxiliary energy conversion device 510 comprises an anolyte compartment 511 and a catholy-te compartment 512 receiving bIeed flows of less than lO~o prefereably less than 5~0 by volume throughou-t, from the recirculating main electrolyte flows, The anolyte of the main cell is catholyte in the aux-iliary cell and the catholyte of -the main cell is anolyte in the auxiliary cell. The auxiliary cell is electrically powered, preferably by trapping the power generation of the main cell to es-tablish an overpotential across the auxiliary cell for overcharging to chlorine potential (2,o volts or higher), Example 6 ;::
Charge-discharge data was plotted as shown in FIG, 6 * Registered Trademark "''~ ' .
., - . .
s~
for a lL~ cell array having :L50 -to 200 m:il thick frames which had feeder holes drilled into edges of -the plates to establish a leak free design. The array was assembled via gaskets and bolted and plates. The anoly-te volume was abou-t hal~ a gallon, catholyte volume about 1.3 gallons and -the molari-ty of the FeC12 solution was 3 to 3 molar. The theoretical maximum capacity of unit as de-termined by 1 3 gal. of 3 2 molar is 1.3 X 4 liters = 5.2 liters of 3.2 molar. This is equivalent to 5 2 X 3.2 = 16.6 moles of FeC12. A-t a maximum o~ 17 ampere-hours per mole, one ob-tains 17 X 16 6 = 283 AH total for the array. Since -there are 14 cells in -the array, each cell could receive 283/14 = 20.2 AH to-tal, if the reagent utiliza-tion factors were lOO~o. No fil-ters or pH adjus-tment apparatus were employed Charging the array a-t 15 amps for 2 hours gave an input of 30 AH. The output was 20 AH at 10 amp rate. Com-plete utilization of electrolyte was realized. Overcharging is evidenced by rise in open circuit voltage and charging voltage towards end of charge mode. Polarization losses are evidenced at this molarity range and dep-th of discharge and curren-t densities by differences in o.c.v. for charge and discharge modes Example 7 The Example 6 14 cell array was operated under same conditions of electrolyte volumes and molarity, etc. as in ~; Example 6, excepting that charging was at 10 amperes and dis-charge at 8 amperes. Results for -two such cycles are shown in FIGS. 7 and 8. The reagent utilization and current efficiency appear quite good, In the FIG, 7 run, there were 5 AH input ..
and almos-t 4 AH output. In the FIG. 8 ru~n, the input was well over 8 AH. Polarizatlon effects, e.g., the differences in - o.c v, of charge and discharge, appear less a-t these lower :
~ 3tj~
curren-t densi-ties compared -to Example 6, Example 8 A 12 cell array assembled in same ~ashion as -the 14 cell array of Examples 6-7, The elec-trolyte was 3 molar FeC12 initially and catholyte volume was about 4 gallons and anolyte 1 gallon. The coulombic capacity of this system is
4 gal. X 4 liters/gal. X 3 molar X 17 AH/mole X 12 cells = 68 AH.
Performance is shown in FIG, 9.
Charging and discharging was done at 12 and 10 amperes, respectively, and this gives some signi~icant polariza-tion losses, A~ter discharging at 10 amperes ~or one hour, the array was placed on open circui-t ~or 20 hours with -the array drained of electrolyte. Corrosion at the anode lowered the potential to 5.~ volts discharge at 10 A when cycling was resumed and energy ef~iciency was consequently reduced.
Example 9 Example 8 was repeated wi-th the di~ference that 4 gallons o~ 3M FeC12 containing 25 gm of sodium citrate per liter as chelating agent was used as catholyte and the resul-tant performance is shown in FIG~. 10 and 11.
Experiments were conducted to learn more about in-ternal resistance changes and polarization voltages as a function o~ current density in a chela-te containing electrolyte.
Resistance values are obtained by dividing current into the difference between open circuit voltage, o.c.v., and cell voltage.
In-ternal resistance lies principally between 0.~ and .
o.6 ohms as measured here. Polarization e~ects do not seem to become pronounced until 10 to 15 amperes drain is reached, Example 10 Then the example 8-9 testing was repeated with no chelate present. The cathol~te was an initial 3M FeC12 , .
:
Performance is shown in FIG, 9.
Charging and discharging was done at 12 and 10 amperes, respectively, and this gives some signi~icant polariza-tion losses, A~ter discharging at 10 amperes ~or one hour, the array was placed on open circui-t ~or 20 hours with -the array drained of electrolyte. Corrosion at the anode lowered the potential to 5.~ volts discharge at 10 A when cycling was resumed and energy ef~iciency was consequently reduced.
Example 9 Example 8 was repeated wi-th the di~ference that 4 gallons o~ 3M FeC12 containing 25 gm of sodium citrate per liter as chelating agent was used as catholyte and the resul-tant performance is shown in FIG~. 10 and 11.
Experiments were conducted to learn more about in-ternal resistance changes and polarization voltages as a function o~ current density in a chela-te containing electrolyte.
Resistance values are obtained by dividing current into the difference between open circuit voltage, o.c.v., and cell voltage.
In-ternal resistance lies principally between 0.~ and .
o.6 ohms as measured here. Polarization e~ects do not seem to become pronounced until 10 to 15 amperes drain is reached, Example 10 Then the example 8-9 testing was repeated with no chelate present. The cathol~te was an initial 3M FeC12 , .
:
5~
solu-tion. Bu-t this array was opera-ted wi-th a pH auxiliary control cell in series (as described above in connection with FIG. 5) and energized at -the same curren-t as -the array. The resultan-t performance is shown in FIG 12. The performance proved to be quite well behaved and with little polarization.
Example 11 A single cell with interelectrode spacing between 0.15 and 0.20 inches was assembled and operated with catholyte comprising one liter of 3.5 M FeC12 solution with one pound o~
ZnC12 dissolved therein and per~ormance is shown in FIG. 13.
The zinc potential, 0.75 volts relative -to hydrogen, was realized However, -the open circuit po~tential decayed rapidly upon discharge to -the iron voltage at the anode, probably because only a limited amount o~ zinc ions are in solution.
Some factor less than lOO~o 0~ the pla-ting ratio o~ Zn -to Fe at the anode during the charging process-is realized. Also, Zn is displaced by Fe 2 ions in the anolyte.
Example 12 The same type o~ experiment as in Example 11 was 20 repeated except that the catholyte was a mixture of 330 ml 9M -~
ZnC12 solution plus 660 ml 4 M FeC12 solution and results are drawn in FIG. 14. The polariza-tion and decay to the Fe/Fe 3 couple voltage is essentially the same as in Example 11.
~; Example 13 FIG 1~ shows the results of a repea-t o~ Example 12 with a catholyte of 600 gm NiC12 salt in a 3.5 molar FeC12 solution. This Ni/Fe 3 cell appears to be well behaved Open circuit potential is less than those o~ the previous examples, as expected, but corrosion and polariza-tion problems are also reduced as shown. Ampere-hour e~iciency is quite high also, about lOO~o according to the curve. Hydrogen gas evolution at .
the Ni plated anode wiIl also be less, as observed in these : :
-tests. We mus-t, however, per~orm more :~u-ture -test~ to fully characterize -this couple and see if the reduced voltage and increased costs as compared to iron are useful trade-offs for -the bet-ter behavior.
Example 13 Sodium citra-te was explored further as solubilizing agents for iron-oxygen compounds, corrosion inhibitor, buffering agents and "brighteners" for iron pla-ting, Resistivi-ty change effects of sodium citra-te, Na3C6H507,5H20, on various molarities of FeC13 solution at 21C are given in -the -table below.
eCl~ Specific Resistivity, ohm-in FeC13 Grams of Sodium Citra~/li-ter Molarit~ 0 10 100 3,42 9.59 -- 9.59 2.89 6.o8 6.36 6,25 2.19 4,82 4,98 4,49 1.58 3.74 -- 3.35 Virtually no effect was observed on specific re-20 sistivity due to the presence of sodium citrate salts up toconcentrations of 100 gm/liter. These small variations in resistance are attributable mainly -to experimental error in weighing, temperature measurements and voltmeter readings.
;General observations made on the effects of citrate on iron plating and sedimentation of electrolytes are:
--Presence of sodium citra-te in electrolytes at a concentration range between 25 and 75 gram/liter very definitely reduces and almost totally prevents the precipita-tion of in- ~-- -soluble iron compounds under normal circumstances. This result , . . .
30 is observed~duri.ng cell operation in the anolyte circui-t and ~
with FeC12 solutions left standing in containers opened to the ~ --alr for prolonged periods of time, ~, .
.
;
~ - 42 -: :
~ ~ 9 3 S~
--Higher concen-trations o~ chela-te, (10 or more grams/li-ter), produce iron pla-ting which has a black appearance as contras-ted with -the usual silvery or metallic finish of iron pla-ting. The plating appears somewha-t porous but quite hard and adhesive. No rapid rusting is observed, --Very high concentrations of citrate, (60 or greater gm/li-ter), seem to produce plating which becomes increasingly spongy and needle-like in character, Plating appears relatively stable in air, --Absence of any chelating agen-t in FeC12 solu-tions produces an iron plating which is no-t qul-te as adhesive -to -the carbon electrode surface and seems -to have a shee-ting or "shingle like" structure which is readily peeled in small sections, When exposed to air, the plating corrodes almost immediately.
--Small quantities of chela-te, (in -the order of 1 gm/liter), has a definite effect on pla-ting, Iron produced from this solu-tion has a relatively smoo-th and silvery appear-ance, is surprisingly stable in air and behaves as a continuous surfacing of metal intimately attached to the carbon electrode substrate.
Based upon investigations, it appears that an initial molarity range of FeC12 solution between 2.5 and l~,o is indicated for optimum cell performance, Chela-ting agent, (sodium ci-trate), concentration of not less than 10 grams/liter and greater than 50 grams/liter -is preferred for best solubilizing results while still retain- `
ing acceptable iron plating characteristics, A single cell of 60 1n2 area electrodes will store 50-60 ampere-hours o~ iron without pronounced iron dendrite shorting, A soIution of 3M FeC12 will provide about 30 to 35 - ampere hours of charge when electrolyzed to a residual concen-~: ." .,:
.
: ~`' .. ' 3~
-tration o~ abou-t o.8M of FeC12. Hence a volume of abou-t 2 li-ters of 3M FeC12 is an appropria-te quan-ti-ty of catholyte for the 4 ampere, 10 hour discharge experirnents.
Barriers should be prewet-ted wi-th a wetting agent, (surfactant), dissolved in me~thanol, e-thanol or hexane for best performance. This need is especially true with the Celgard reinforced barrier which is employed -throughou-c several of the units described above.
Electrolyte flow rates should not be less than 2cc per second on -those cells with 0.10 in in-terelectrode spacing Such flows appear to be adequa-te supply rates of ionic species at the electrode surfaces with current densities up to 0 20 amp/in2, Cells with spacing of 0.10 in. between electrodes do no-t, unfortunately, make maximum use of -the low electrolyte resis-tance. Resistance checks with full scale cells show that most of the resistance is the result of electrode properties.
A cell with 60 in2 area filled with 2 8 molar solution of FeC12 and no barrier has a total resistance, at 24C, of 0 016 ohms About 0.004 ohms is due to the elec-trolyte and somewhere in the region of 0.008 and 0 012 ohms are contribu-ted by ~the electrode, depending upon concen-tration of electrolyte and wetting agents present. If a plain Celgard* barrier is employed in such a ;
cell its total resistance will be about 0 016 ohms with only 1/3 of this quantity represented by -the electrolytes A
reinforced Celgard, prewetted, barrier in the same cell results in a total resistance of 0 020 ohms. About 1/4 of the total is now represented by the elec-trolyte. ~
The use of chelation involves sequestration of iron, -and/or other pla-ting/deplating metal(s) used, to prevent ':
* Registered Trademark oxidation and aid pla-ting and redissolu-tion ~thereof, In particular Fe 2 ions are preferrentially chelated -to make -them unavailable for -the reaction Fe 2 + 2e >Fe to enhance e~ficiency of zinc plating (where zinc and iron salts are both dissolved in the anolyte). ~helating agen-ts particularly useful for the last mentioned purpose comprise CX ~ / - dipyrridyl O - phenanthroline protoporphysin nitro - O - phenanthroline As inhibitors of oxygen compounds of iron, one may use citrates, -tartarates, oxala-tes, ace-ta-tes, amines, pyrrolidine, ammonium sulfate, fluorides, thiocyanates and ether chains, In addition to or in lieu of pH control to avoid the formation of insoluble iron compounds, the invention comprehends excluding air from the system by hermetic sealing and utilizing a non-wa-ter solvent for the salt. Alcohol is one such possibility. No H or OH- ions are formed and -there are no electrolytic decomposition products of methanol or ethanol to react with ironO FeC12, (anhydrous), is soluble in alcohol to about 100 gm of salt per 100 ml of solvent. FeC13 is also very soluble in alcohol, A further important aspect of the invention con-trol of the cathode surface, (surface of -the positive electrode), which must have some charac-teristics which will cause the reactions to proceed with minimum starvation and poisoning or floodlng effects" The following electrochemical half cell reaction occurs, During charge - oxidation of Fe 2, (ferrous ions), . .
proceeds, ; ~ Fe 2 ~ Fe 3 + e During discharge - reduction of the Fe 3, (ferric - 4~ _ ions), -takes place, Fe~3 t e --~Fe+2 During ei-ther of -these processes no iron ions, Fe 2 or Fe 3, are a-t-tracted -to this electrode surface as a result of electric field gradien-ts or electric polari-ty attraction because -the electrode, (ca-thode), is (-~) as are -the ionic charges. Instead chloride ions are caused to migrate -to -the cathode via electric fields and currents, FIGS, 16A and 16~
graphically illustra-te -this si-tua-tion for charge and discharge processes, respec-tively, Since the iron ions are no-t supplied to the electrode surface where -the oxida-tion or reduction pro-cess takes place, they mus-t be supplied "mechanically", "physically" e,g,, via forced fluid flow and thermal or molecular diffusion, Hence elec-trolyte is caused to flow near and over the surfacé of the cathode to proviae for good mixing and supply of reagents. The electrode surface must have the following properties, --high electronic conductivity --low work function for the oxidation - reduction electron exchange process, --chemical inertness, ---easily manufacturable and attachable to -the conductive substrate.
--low cos-t, readily available materials.
--physical structure which optimizes the necessary transport mechanisms, The surface area of -the cathode should no-t only be as large as possible per frontal area of electrode but i-t must also be available or accessible to the inflow of Fe 2 ions and exhaust flow of Fe 3 ions during the charging mode and the opposite flow pattern during discharge, .:
. -An effec-tive cathode surf'ace in -terms of such criteria is achieved wi-th 40-80 mesh OL charcoal, manufac-tured by the Calgon Corp. bonded -to the surface of a graphite-plastic con-ductive substrate. Polarization voltage drops less -than 0.01 to 0, 02 vol-ts are obtained with such cathodes all the way up to current densities of l amp/in2 with cells 80 to 90% dis-charged or with catholyte concentrations in the order of 0. 5 to l.O M FeC13, Substrate composition is also importan-t following compositions of substrates (for anode and cathode) subs~ti-tuting polyvinyl chloride for more expensive fluoro-hydrocarbon binders allow cost reduc-tion and their resis-tivi-ties are as follows:
Dixon Graphite ~1112 f-- 30% by weigh-t Diamond Shamrock PVC resin ~5r~7~ by weigh-t Bulk resistivity in direction of plane of sneet f = o o25 ohm-in.
. .
Dixon Graphite ~8485 ~ 70%
Diamond Shamrock PVC ~71AH ~ 30%
p = o. o53 ohm-in.
Graphite ~8485 60% ~ :
B,F. Goodrich PVC #124 40%
= o, o7 ohm-in.
Graphite #8485 30%
Goodrich ~124 70% ~ :
= 0.29 ohm-in.
Graphite ~8485 70%
Goodrich PVC #126 30~o - ~ = 0,11 ohm-in.
~: .
.
. : .
1~'7~
Graphite ~8485 5 Goodrich PVC ~124 50%
~ = 0,13 ohm-in.
These plates are pressed in a con-tained mold with about 100 in area under -the following conditions.
Total force ~ 15 tons Temperature f~ 350~
Pressing time ~ 10 minutes The dry powders are mixed in a PK blender for 15 minutes with rotating mixing bar.
These materials all appear qui-te s-trong mechanically and potentially sui-table for the fuel cell system. A maximum resis-tivity value in -the range of 0.10 ohm-in appears to be practical for substrate resistivity with -the following criteria:
--Electrodes should not contribute more than 10% to the to-tal resistance of a cell.
--Electrodes may be in the range of 0.05 inches thick, --Volume resis-tivi-ty of graphite composites may be 5 to 10 times greater in the direction normal to the sheet than that value measured in the plane of the plate.
--Electrolyte resistivity will be in the order of 2 to 6 ohms-in.
In any of the foregoing embodiments of -the invention state of charge may be sensed by the absolute direct reading system of FIG, 17A or the compara-tive readout system of FIG. 17B in both which a (color) light source is ~S and a filter .
F pass through electrolyte in absorption cell(s) AS (after reflecting off beam splltter mirror BS in -the FIG. 17B case) and then~to photo-cell~s) PC and an amplifier A and an amplifier A
and readout on meter M or passage of a state of charge ~ -~
correlated electrical signal to an automatic controller not ~, :: :
~ ~ - 48 -7~3~
shown. The systems are calibra-ted ln accordance with Beer's ~aw I = Io e X
where Io is intensity o~ inciden-t ligh-t on cell AS, I is transmitted light intensity, X is absorp-tion cell -thickness and ~ is -the electroly-te absortivity which is calibratable -to con-centration of dissolved solute (and more particularly di~er-entiating Fe 2 and Fe 3 in connection wi-th such concentration measurements) and thus -to s-tate of charge (or discharge).
Such state of charge measurement may also be carried out electrically by measuring elec-trolyte po-tential wi-th re-spect to a reference elec-trode.
It is eviden-t that -those skilled in ~the art, once given the benefit of -the foregoing disclosure, may now make numerous other uses and modi~ications of~ and departures from ~ .
the specific embodimen-ts described herein wi-thout departing from the inventive concepts Consequently, the inven-tion is to be construed as embracing each and every novel feature and ~ .
novel combination o~ features present in, or possessed by, ::~
20 the apparatus and techniques herein disclosed and limited ~ .
solely oy the scope and spirit o~ the appended claims.
: ' .
~ 49 ~
solu-tion. Bu-t this array was opera-ted wi-th a pH auxiliary control cell in series (as described above in connection with FIG. 5) and energized at -the same curren-t as -the array. The resultan-t performance is shown in FIG 12. The performance proved to be quite well behaved and with little polarization.
Example 11 A single cell with interelectrode spacing between 0.15 and 0.20 inches was assembled and operated with catholyte comprising one liter of 3.5 M FeC12 solution with one pound o~
ZnC12 dissolved therein and per~ormance is shown in FIG. 13.
The zinc potential, 0.75 volts relative -to hydrogen, was realized However, -the open circuit po~tential decayed rapidly upon discharge to -the iron voltage at the anode, probably because only a limited amount o~ zinc ions are in solution.
Some factor less than lOO~o 0~ the pla-ting ratio o~ Zn -to Fe at the anode during the charging process-is realized. Also, Zn is displaced by Fe 2 ions in the anolyte.
Example 12 The same type o~ experiment as in Example 11 was 20 repeated except that the catholyte was a mixture of 330 ml 9M -~
ZnC12 solution plus 660 ml 4 M FeC12 solution and results are drawn in FIG. 14. The polariza-tion and decay to the Fe/Fe 3 couple voltage is essentially the same as in Example 11.
~; Example 13 FIG 1~ shows the results of a repea-t o~ Example 12 with a catholyte of 600 gm NiC12 salt in a 3.5 molar FeC12 solution. This Ni/Fe 3 cell appears to be well behaved Open circuit potential is less than those o~ the previous examples, as expected, but corrosion and polariza-tion problems are also reduced as shown. Ampere-hour e~iciency is quite high also, about lOO~o according to the curve. Hydrogen gas evolution at .
the Ni plated anode wiIl also be less, as observed in these : :
-tests. We mus-t, however, per~orm more :~u-ture -test~ to fully characterize -this couple and see if the reduced voltage and increased costs as compared to iron are useful trade-offs for -the bet-ter behavior.
Example 13 Sodium citra-te was explored further as solubilizing agents for iron-oxygen compounds, corrosion inhibitor, buffering agents and "brighteners" for iron pla-ting, Resistivi-ty change effects of sodium citra-te, Na3C6H507,5H20, on various molarities of FeC13 solution at 21C are given in -the -table below.
eCl~ Specific Resistivity, ohm-in FeC13 Grams of Sodium Citra~/li-ter Molarit~ 0 10 100 3,42 9.59 -- 9.59 2.89 6.o8 6.36 6,25 2.19 4,82 4,98 4,49 1.58 3.74 -- 3.35 Virtually no effect was observed on specific re-20 sistivity due to the presence of sodium citrate salts up toconcentrations of 100 gm/liter. These small variations in resistance are attributable mainly -to experimental error in weighing, temperature measurements and voltmeter readings.
;General observations made on the effects of citrate on iron plating and sedimentation of electrolytes are:
--Presence of sodium citra-te in electrolytes at a concentration range between 25 and 75 gram/liter very definitely reduces and almost totally prevents the precipita-tion of in- ~-- -soluble iron compounds under normal circumstances. This result , . . .
30 is observed~duri.ng cell operation in the anolyte circui-t and ~
with FeC12 solutions left standing in containers opened to the ~ --alr for prolonged periods of time, ~, .
.
;
~ - 42 -: :
~ ~ 9 3 S~
--Higher concen-trations o~ chela-te, (10 or more grams/li-ter), produce iron pla-ting which has a black appearance as contras-ted with -the usual silvery or metallic finish of iron pla-ting. The plating appears somewha-t porous but quite hard and adhesive. No rapid rusting is observed, --Very high concentrations of citrate, (60 or greater gm/li-ter), seem to produce plating which becomes increasingly spongy and needle-like in character, Plating appears relatively stable in air, --Absence of any chelating agen-t in FeC12 solu-tions produces an iron plating which is no-t qul-te as adhesive -to -the carbon electrode surface and seems -to have a shee-ting or "shingle like" structure which is readily peeled in small sections, When exposed to air, the plating corrodes almost immediately.
--Small quantities of chela-te, (in -the order of 1 gm/liter), has a definite effect on pla-ting, Iron produced from this solu-tion has a relatively smoo-th and silvery appear-ance, is surprisingly stable in air and behaves as a continuous surfacing of metal intimately attached to the carbon electrode substrate.
Based upon investigations, it appears that an initial molarity range of FeC12 solution between 2.5 and l~,o is indicated for optimum cell performance, Chela-ting agent, (sodium ci-trate), concentration of not less than 10 grams/liter and greater than 50 grams/liter -is preferred for best solubilizing results while still retain- `
ing acceptable iron plating characteristics, A single cell of 60 1n2 area electrodes will store 50-60 ampere-hours o~ iron without pronounced iron dendrite shorting, A soIution of 3M FeC12 will provide about 30 to 35 - ampere hours of charge when electrolyzed to a residual concen-~: ." .,:
.
: ~`' .. ' 3~
-tration o~ abou-t o.8M of FeC12. Hence a volume of abou-t 2 li-ters of 3M FeC12 is an appropria-te quan-ti-ty of catholyte for the 4 ampere, 10 hour discharge experirnents.
Barriers should be prewet-ted wi-th a wetting agent, (surfactant), dissolved in me~thanol, e-thanol or hexane for best performance. This need is especially true with the Celgard reinforced barrier which is employed -throughou-c several of the units described above.
Electrolyte flow rates should not be less than 2cc per second on -those cells with 0.10 in in-terelectrode spacing Such flows appear to be adequa-te supply rates of ionic species at the electrode surfaces with current densities up to 0 20 amp/in2, Cells with spacing of 0.10 in. between electrodes do no-t, unfortunately, make maximum use of -the low electrolyte resis-tance. Resistance checks with full scale cells show that most of the resistance is the result of electrode properties.
A cell with 60 in2 area filled with 2 8 molar solution of FeC12 and no barrier has a total resistance, at 24C, of 0 016 ohms About 0.004 ohms is due to the elec-trolyte and somewhere in the region of 0.008 and 0 012 ohms are contribu-ted by ~the electrode, depending upon concen-tration of electrolyte and wetting agents present. If a plain Celgard* barrier is employed in such a ;
cell its total resistance will be about 0 016 ohms with only 1/3 of this quantity represented by -the electrolytes A
reinforced Celgard, prewetted, barrier in the same cell results in a total resistance of 0 020 ohms. About 1/4 of the total is now represented by the elec-trolyte. ~
The use of chelation involves sequestration of iron, -and/or other pla-ting/deplating metal(s) used, to prevent ':
* Registered Trademark oxidation and aid pla-ting and redissolu-tion ~thereof, In particular Fe 2 ions are preferrentially chelated -to make -them unavailable for -the reaction Fe 2 + 2e >Fe to enhance e~ficiency of zinc plating (where zinc and iron salts are both dissolved in the anolyte). ~helating agen-ts particularly useful for the last mentioned purpose comprise CX ~ / - dipyrridyl O - phenanthroline protoporphysin nitro - O - phenanthroline As inhibitors of oxygen compounds of iron, one may use citrates, -tartarates, oxala-tes, ace-ta-tes, amines, pyrrolidine, ammonium sulfate, fluorides, thiocyanates and ether chains, In addition to or in lieu of pH control to avoid the formation of insoluble iron compounds, the invention comprehends excluding air from the system by hermetic sealing and utilizing a non-wa-ter solvent for the salt. Alcohol is one such possibility. No H or OH- ions are formed and -there are no electrolytic decomposition products of methanol or ethanol to react with ironO FeC12, (anhydrous), is soluble in alcohol to about 100 gm of salt per 100 ml of solvent. FeC13 is also very soluble in alcohol, A further important aspect of the invention con-trol of the cathode surface, (surface of -the positive electrode), which must have some charac-teristics which will cause the reactions to proceed with minimum starvation and poisoning or floodlng effects" The following electrochemical half cell reaction occurs, During charge - oxidation of Fe 2, (ferrous ions), . .
proceeds, ; ~ Fe 2 ~ Fe 3 + e During discharge - reduction of the Fe 3, (ferric - 4~ _ ions), -takes place, Fe~3 t e --~Fe+2 During ei-ther of -these processes no iron ions, Fe 2 or Fe 3, are a-t-tracted -to this electrode surface as a result of electric field gradien-ts or electric polari-ty attraction because -the electrode, (ca-thode), is (-~) as are -the ionic charges. Instead chloride ions are caused to migrate -to -the cathode via electric fields and currents, FIGS, 16A and 16~
graphically illustra-te -this si-tua-tion for charge and discharge processes, respec-tively, Since the iron ions are no-t supplied to the electrode surface where -the oxida-tion or reduction pro-cess takes place, they mus-t be supplied "mechanically", "physically" e,g,, via forced fluid flow and thermal or molecular diffusion, Hence elec-trolyte is caused to flow near and over the surfacé of the cathode to proviae for good mixing and supply of reagents. The electrode surface must have the following properties, --high electronic conductivity --low work function for the oxidation - reduction electron exchange process, --chemical inertness, ---easily manufacturable and attachable to -the conductive substrate.
--low cos-t, readily available materials.
--physical structure which optimizes the necessary transport mechanisms, The surface area of -the cathode should no-t only be as large as possible per frontal area of electrode but i-t must also be available or accessible to the inflow of Fe 2 ions and exhaust flow of Fe 3 ions during the charging mode and the opposite flow pattern during discharge, .:
. -An effec-tive cathode surf'ace in -terms of such criteria is achieved wi-th 40-80 mesh OL charcoal, manufac-tured by the Calgon Corp. bonded -to the surface of a graphite-plastic con-ductive substrate. Polarization voltage drops less -than 0.01 to 0, 02 vol-ts are obtained with such cathodes all the way up to current densities of l amp/in2 with cells 80 to 90% dis-charged or with catholyte concentrations in the order of 0. 5 to l.O M FeC13, Substrate composition is also importan-t following compositions of substrates (for anode and cathode) subs~ti-tuting polyvinyl chloride for more expensive fluoro-hydrocarbon binders allow cost reduc-tion and their resis-tivi-ties are as follows:
Dixon Graphite ~1112 f-- 30% by weigh-t Diamond Shamrock PVC resin ~5r~7~ by weigh-t Bulk resistivity in direction of plane of sneet f = o o25 ohm-in.
. .
Dixon Graphite ~8485 ~ 70%
Diamond Shamrock PVC ~71AH ~ 30%
p = o. o53 ohm-in.
Graphite ~8485 60% ~ :
B,F. Goodrich PVC #124 40%
= o, o7 ohm-in.
Graphite #8485 30%
Goodrich ~124 70% ~ :
= 0.29 ohm-in.
Graphite ~8485 70%
Goodrich PVC #126 30~o - ~ = 0,11 ohm-in.
~: .
.
. : .
1~'7~
Graphite ~8485 5 Goodrich PVC ~124 50%
~ = 0,13 ohm-in.
These plates are pressed in a con-tained mold with about 100 in area under -the following conditions.
Total force ~ 15 tons Temperature f~ 350~
Pressing time ~ 10 minutes The dry powders are mixed in a PK blender for 15 minutes with rotating mixing bar.
These materials all appear qui-te s-trong mechanically and potentially sui-table for the fuel cell system. A maximum resis-tivity value in -the range of 0.10 ohm-in appears to be practical for substrate resistivity with -the following criteria:
--Electrodes should not contribute more than 10% to the to-tal resistance of a cell.
--Electrodes may be in the range of 0.05 inches thick, --Volume resis-tivi-ty of graphite composites may be 5 to 10 times greater in the direction normal to the sheet than that value measured in the plane of the plate.
--Electrolyte resistivity will be in the order of 2 to 6 ohms-in.
In any of the foregoing embodiments of -the invention state of charge may be sensed by the absolute direct reading system of FIG, 17A or the compara-tive readout system of FIG. 17B in both which a (color) light source is ~S and a filter .
F pass through electrolyte in absorption cell(s) AS (after reflecting off beam splltter mirror BS in -the FIG. 17B case) and then~to photo-cell~s) PC and an amplifier A and an amplifier A
and readout on meter M or passage of a state of charge ~ -~
correlated electrical signal to an automatic controller not ~, :: :
~ ~ - 48 -7~3~
shown. The systems are calibra-ted ln accordance with Beer's ~aw I = Io e X
where Io is intensity o~ inciden-t ligh-t on cell AS, I is transmitted light intensity, X is absorp-tion cell -thickness and ~ is -the electroly-te absortivity which is calibratable -to con-centration of dissolved solute (and more particularly di~er-entiating Fe 2 and Fe 3 in connection wi-th such concentration measurements) and thus -to s-tate of charge (or discharge).
Such state of charge measurement may also be carried out electrically by measuring elec-trolyte po-tential wi-th re-spect to a reference elec-trode.
It is eviden-t that -those skilled in ~the art, once given the benefit of -the foregoing disclosure, may now make numerous other uses and modi~ications of~ and departures from ~ .
the specific embodimen-ts described herein wi-thout departing from the inventive concepts Consequently, the inven-tion is to be construed as embracing each and every novel feature and ~ .
novel combination o~ features present in, or possessed by, ::~
20 the apparatus and techniques herein disclosed and limited ~ .
solely oy the scope and spirit o~ the appended claims.
: ' .
~ 49 ~
Claims (13)
1. Energy conversion apparatus comprising, means defining at least a single electrolytic cell comprising anolyte and catholyte compartments separated by a diffusion barrier, means defining anode and cathode plastic carbon com-posite respectively in said anolyte and catholyte compartment, means defining anolyte and catholyte liquids re-circulating in closed loops which contain said compartments respectively, and recirculating and bulk storage means for said fluids, the fluids being solutions of metal salt which plate out metal on the anode during application of charging potential to the electrodes and deplate metal and redissolve it into the anolyte during discharge of the system and which change valence states within the catholyte without plating, electrolytic means for maintaining catholyte molarity between 2.0 and 4.0, means for preventing oxidation of the plating/de-plating metal to combat polarization and maintain reversibility over extended charge/discharge cycling.
2. Energy conversion apparatus in accordance with claim 1 and further comprising, means defining an auxiliary electrolytic device with at least a single cell pair of anolyte and catholyte compart-ments separated by diffusion barrier means and comprising therein anode and cathode electrode means, respectively, and a power supply connected to said anode and cathode electrodes for maintaining across said electrodes an overvoltage (compared to the full charge potential of the electrochemical pair of the main electrolytic device), means for bleeding a minor portion of the recirculat-ing anolyte of the main electrolytic device through said auxiliary electrolytic device and returning it to the recircu-lating loop of the main electrolytic device, the auxiliary electrolytic device being poled so that said anolyte from the main electrolytic device passes through the catholyte compartment of the auxiliary electrolytic device and so that pH of main electrolytic device anolyte is decreased by electrolytic production of hydrogen ion the auxiliary electrolytic device to counteract the pH-increasing tendency of the main electrolytic device, whereby stable operation at high efficiency is enabled over long periods of time, and means for essentially continuously passing an electrolyte through the anolyte compartment of the auxiliary device.
3. Energy conversion apparatus in accordance with claim 2 wherein said means for continuously passing electrolyte through the anolyte compartment of the auxiliary electrolytic device comprises a bleed from, and return to, the recirculating catholyte, of the main electrolytic device, of a minor portion of the said catholyte.
4. Energy conversion apparatus in accordance with claim 3 wherein said bleed electrolyte flows comprise less than 10%
of main flows of the respective electrolytes from which they are bled.
of main flows of the respective electrolytes from which they are bled.
5. Energy conversion apparatus in accordance with claim 1 wherein said anolyte and catholyte comprise at least one cation selected from the group consisting of iron, nickel and zinc and a chloride anion.
6. Energy conversion apparatus in accordance with claim 5 wherein said anolyte comprises iron and zinc components which plate out as an alloy layer on the anode in charging and substantially reversibly redissolve in the anolyte on dis-charge.
7. Energy conversion apparatus in accordance with claim 1, and further comprising means for draining catholyte from said catholyte compartments back to its reservoir to turn off power generation.
8. Energy conversion apparatus in accordance with claim 1 and further comprising, means defining filters in the recirculating anolyte and catholyte loops, and means for periodically switching the filters between loops to allow chemical cleansing of such filters.
9. Energy conversion apparatus in accordance with claim 1 comprising an aqueous solvent in said anolyte solution.
10. Energy conversion apparatus in accordance with claim 1 comprising a non-aqueous solvent in said anolyte solution.
11. Energy conversion apparatus in accordance with claim 1 comprising a chelating agent in said anolyte solution to sequester the plating/deplating metal thereby enhancing plating and redissolution thereof and preventing its oxidation.
12. Energy conversion apparatus in accordance with claim 1 comprising a multi-cell array.
13. Energy conversion apparatus in accordance with claim 12 wherein individual cells of the array are arranged back-to-back using bipolar electrodes.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CA291,579A CA1079350A (en) | 1977-11-23 | 1977-11-23 | Fuel cell using iron chloride electrolytes |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CA291,579A CA1079350A (en) | 1977-11-23 | 1977-11-23 | Fuel cell using iron chloride electrolytes |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CA1079350A true CA1079350A (en) | 1980-06-10 |
Family
ID=4110116
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CA291,579A Expired CA1079350A (en) | 1977-11-23 | 1977-11-23 | Fuel cell using iron chloride electrolytes |
Country Status (1)
| Country | Link |
|---|---|
| CA (1) | CA1079350A (en) |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN106414809A (en) * | 2013-12-16 | 2017-02-15 | 阿海珐 | Large-capacity electrical-energy storage device |
-
1977
- 1977-11-23 CA CA291,579A patent/CA1079350A/en not_active Expired
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN106414809A (en) * | 2013-12-16 | 2017-02-15 | 阿海珐 | Large-capacity electrical-energy storage device |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| US4069371A (en) | Energy conversion | |
| Vincent et al. | Modern batteries | |
| US8158277B1 (en) | Cross-flow electrochemical batteries | |
| CN102110839A (en) | Battery | |
| DE202011051993U1 (en) | Multi-mode loading of hierarchical anode | |
| BG61627B1 (en) | Method and system forelectrochemical energyaccumulation and/or power generation | |
| GB1569398A (en) | Metal halogen batteries and method of operating same | |
| JPH08502386A (en) | Electrochemical device for power distribution using air electrode | |
| US3909298A (en) | Batteries comprising vented electrodes and method of using same | |
| EP0109807A1 (en) | Zinc-bromine battery with long term stability | |
| Pichler et al. | The impact of operating conditions on component and electrode development for zinc-air flow batteries | |
| EP1371067B1 (en) | Compound having a high conductivity for electrons; electrode for an electrochemical cell which comprises this compound, method for preparing an electrode and electrochemical cell | |
| CN101636859A (en) | Mesoporous electrodes for electrochemical cells | |
| Gahn et al. | Single cell performance studies on the Fe/Cr Redox Energy Storage System using mixed reactant solutions at elevated temperature | |
| MU¨ LLER et al. | Development of a 100 W rechargeable bipolar zinc/oxygen battery | |
| JP3315508B2 (en) | Electrolyte flow battery with electrolyte readjustment device | |
| CA1079350A (en) | Fuel cell using iron chloride electrolytes | |
| EP3565048A1 (en) | Redox flow battery | |
| Munshi et al. | The dissolution of iron from the negative material in pocket plate nickel-cadmium batteries | |
| Lu et al. | Introduction to Zinc–Air Batteries | |
| GB1587748A (en) | Energy conversion apparatus | |
| EP2001072A2 (en) | Powdered Fuel Cell | |
| Yuan et al. | Hail to Daniell Cell: From Electrometallurgy to Electrochemical Energy Storage | |
| Kazacos | Electrolyte optimization and electrode material evaluation for the vanadium redox battery | |
| DE2129045C3 (en) | Electrochemical cell for energy storage in which the electrode of one polarity is a battery electrode, the other polarity is a gas electrode |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| MKEX | Expiry |