Overview:

This research was conducted in four phases:

• 1998-1999 : Development of Lithium cell design and testing varied cathode metal-salts. Electrolyte selected.
• 2000-2001 : Testing various concentrations of electrolyte on current output over 400s interval.
• 2001-2002 : Testing of various anions in lithium cells.
• 2002-2003 : Development of ion gradient theory, expanded data logging and more controlled environment allows automated logging for 3600seconds and multiple trials. Research in water concentration affecting gradient formation and tabulations of peak and steady state current. Research presented at Intel ISEF.

Read the paper here , or on the web below.

Abstract

The requirement of an anhydrous and aprotic electrolyte in lithium cells prevents the use of aqueous electrolytes with unity transference. The use of non-unity transference electrolytes causes concentration gradients to form reducing the current output of the cell by increasing electrolyte resistance.
These experiments were designed to analyze the performance of primary lithium cells when the radii of the anion, the concentration of the electrolyte, and the amount of water contamination in the electrolyte were varied. This effects the rate of gradient formation and the magnitude of the gradient at steady-state current.
It was concluded that water plays a role in increasing the rate of gradient formation of ions with smaller radii. Furthermore, it was concluded that a high electrolyte concentration increases the formation rate of gradients by decreasing ion mobility due to a high viscosity.

Lithium cells are becoming an increasingly important part of portable power sources due to their high energy density and low weight. To fulfill their full potential, lithium cells must overcome several limitations. The most significant of these limitations is the formation of ion gradients in the cell’s electrolyte. Such gradients impede the amount of current that the cell can produce. Overcoming this limitation will allow lithium cells to output an increased amount of current.

Based on earlier studies that have been conducted on lithium / silver salt cells, it was observed that current output depended on electrolyte concentration, viscosity, and the particular anion of the silver salt. From reading and consultation, I have come to focus subsequent studies on the effect of these variables on the formation of ion gradients in these cells as an underlying factor effecting the current output.

Therefore it is hypothesized that Li | Ag-salt cells utilizing silver salts with anions of increasing radii will form ion gradients of different magnitudes. These gradients will be observable as the peak in current output. It is further hypothesized that Li | Ag-salt cells will have decreasing ion transference rates in a LiBF4 electrolyte of increasing concentration when the anion is not varied. This will lead to the formation of ion gradients of increasing magnitude. The experiments described in this report were intended to test the above stated hypotheses. Furthermore, the results of the presented research will determine the optimum ion radii and electrolyte concentration for lithium cells in order to optimize output current.

Concentration Gradient formation has been a known problem in lithium cells. Concentration gradients form when the rate of ion generation is not matched in rate by ion transfer across the cell. Ion diffusion is driven by entropy, as the ions try to become evenly distributed across the cell. Initially, when current begins to flow through a cell, there are no ion gradients present, the electrolyte resistance is at a minimum, and therefore the cell is producing the maximum amount of current possible; this is called initial current (I0). During current generation, an ion gradient forms in a lithium cell since the rate of ion transport away from the anode and cathode of the cell is less than the rate at which the ions accumulate at the originating electrode.

Once ion gradients form, ion transport becomes limited and the resistance of the electrolyte increases. This increase in electrolyte resistance is due to concentration polarization of the electrolyte is defined as a build up of anion/cation in a cell’s electrolyte. Concentration polarization reduces the output current by introducing a counter EMF in the electrolyte of a cell, therefore increasing the cell’s internal resistance to the flow of current. This reduces the output current until the rate of ion generation is equivalent to the rate of ion transport. When this equilibrium point is reached, the current output remains constant; this is called steady state current.

An ion’s radius has an effect on gradient formation in Li cells. Therefore, ion radii calculated using the Kapustinskii equation [1] were compared to current output. As anionic radii increase, the volume of an anion increases, therefore distributing the anionic charge over a greater volume. This decrease in the anion’s charge density has an effect on intermolecular forces between an anion and water contamination in the electrolyte.

The current output of a lithium cell is also dependent on the electrolyte used. Since "lithium is more electropositive than hydrogen, the electrolyte must be nonaqueous and aprotic" [2]. This requirement prevents the use of aqueous electrolytes with unity transference in lithium cells. The use of non-unity transference electrolytes causes formation of concentration gradients, reducing the current output of the cell by increasing electrolyte resistance. Experiments described in this report were designed to analyze the performance of primary lithium cells when the concentration of the electrolyte used in the cell was varied for a given silver salt. It was postulated that this would effect the rate of gradient formation, the magnitude of the gradient, and current output when steady-state current is achieved.

Together, the studies described here were carried out to determine the effects that anion size, and the corresponding anionic charge density, and electrolyte concentrations have on gradient formation and current output in Li cells.

Materials

1tank Argon
1 Dryrite tube
1 Lab. Con. Co. 50004 glove box
1sheet Bounty two-ply Paper towels
1 Wire cutter
1 razorblade
45 1ml plastic Tuberculin syringes
1 Craftsman 82325 PC interface multimeter
1 Glue gun
4 Hot melt glue sticks
35cm 3.4 mm Lithium wire (1% Sodium)
100ml C4H10O2 1,2-Dimethoxy ethane, CAS: 110-71-4, 99.5% Anhydrous
100ml C4H6O3 Propylene carbonate, CAS: 108-32-7, 99.7% Anhydrous
10g LiBF4 Lithium tetrafluoroborate, CAS: 14283-07-9, 98% Anhydrous
5g AgF Silver Fluoride, CAS: 7775-41-9
5g AgCN Silver Cyanide, CAS: 506-64-9
5g AgI Silver Iodide, CAS: 7783-96-2
5g AgBr Silver Bromide, CAS: 7785-23-1
5g AgNO3 Silver Nitrate, CAS: 7761-88-8
5g AgClO4 Silver Perchlorate, CAS: 7783-93-9
1sheet 160# sandpaper
10ml Hexane
6 20 ml scintillation vials
6 1dram vials
6 6cmX6cm-aluminum foil squares
1 25 ml volumetric flask
2 18" 16 gauge non-coring needle
4m 18 AWG bare copper wire
1 Tweezers

1. Preparation of chemicals
1a. Prepare a 0.0 g/ml stock solution of [electrolyte LiBF4] in 1,2-Dimethoxyethane and Propylene carbonate; 1:1 by volume. .
1b. Prepare stock solution of 0.4 g/ml LiBF4 in 1,2-Dimethoxyethane: Propylene carbonate, 1:1 by volume, of electrolyte
1c. Similarly, prepare the following electrolyte concentrations: 0.0 g/ml, 0.1 g/ml, 0.2 g/ml, 0.3 g/ml, and 0.4 g/ml .
1d. Grind each of the 6 silver salts listed in the “Materials” section to a fine powder (150 mesh) and add them to separate 1dram vials wrapped in aluminum foil.

2. Construction of cells
2a. Cut 1cm long pieces of the 3.4 mm lithium wire into a scintillation vial filled with hexane.
2b. Remove the plungers from 1 ml Tuberculin syringes and cut the barrels at the 0.6ml mark with a razorblade.
2c. Sand the 18 AWG un-insulated copper wire using grade 160# sandpaper.
2d. Cut 55 mm pieces from the copper wire and construct cathode contacts per Fig 3-1.

2e. Construct anode contacts per Fig 3-2.

2f. Insert the anode contact, long end first, through the syringe barrel into the tip end of the syringe so that the straight piece of wire extends through the end of the syringe and the loop end is in the inside of the syringe barrel. Push the copper wire in until the loop reaches the bottom. Connect the barrel to a vacuum source and draw hot melt glue into the tip until it reaches the bottom of the loop.
2g. To construct separators, cut 3 cm X 2 cm rectangular pieces out of the Bounty paper towel. Fold them into thirds along the 3 cm direction and then roll the folded towel into a cylinder. Insert 10 separators into a syringe barrel from which the tip was cut off. This will allow the separators to be pushed out into the cells one at a time.
2h. Assemble lithium cells per Fig 3-3.

Note: Cells were tested in a vertical position with the cathode up.
Testing Li-cell properties
The lithium cells were assembled per Fig 3-3. Cells were tested in a vertical position with the cathode up.
(1) Effect of electrolyte concentration on current output
1a. Insert a cleaned 1 cm piece of lithium into the syringe barrel and compact the wire so it forms a tight seal at the edges of the syringe and is in electrical contact with the copper anode contact.
1b. Add 0.3 ml of a variable concentration electrolyte, (0.0 g/ml, 0.1 g/ml, 0.2 g/ml, 0.3 g/ml, or 0.4 g/ml).
1c. Inject one separator into the lithium cell. Subject cell to light vacuum to remove argon bubbles from separator.
1d. Start timer and add 0.2 ml of AgNO3.
1e. After 90 sec insert the cathode contact and connect multimeter.
1f. After 130 sec start data logging. Logging program will record current output of the cell every second over a duration of 400 seconds.
1g. Repeat for electrolyte concentrations 0.0 g/ml through 0.4 g/ml

(2) Effect of electrolyte concentration on electrolyte conductivity
2a. Add 0.4 ml of variable concentration electrolyte (0.0 g/ml, 0.1 g/ml, 0.2 g/ml, 0.3 g/ml, or 0.4 g/ml) into a syringe barrel.
2b. Insert the cathode contact 1cm from anode contact and connect multimeter.
2c. Start data logging. Logging program will record resistance of the cell every second for a duration of 400 seconds.
2d. Repeat for electrolyte concentrations 0.0 g/ml through 0.4 g/ml.

(3) Effect of electrolyte concentration on electrolyte viscosity.
3a. Measure electrolyte fall time through a 1 ml TD pipette from the 0.9 ml mark to the 0.0 ml mark (viscosometer, Fig 3-4) for electrolytes (0.0 g/ml, 0.1 g/ml, 0.2 g/ml, 0.3 g/ml, or 0.4 g/ml).
3b Calibrate viscosometer by measuring the fall time of water from the 0.9 ml mark to the 0.0 ml mark.

(4) Effect of ion radius on gradient formation and steady state current
4a. Insert a cleaned 1 cm piece of lithium into the syringe barrel and compact the wire so it forms a tight seal at the edges of the syringe and is in electrical contact with the copper anode contact.
4b. Add 0.3 ml of the 0.2 g/ml electrolyte solution.
4c. Inject one separator into the lithium cell. Subject cell to light vacuum to remove argon bubbles from separator.
4d. Start timer and add 0.2 ml of cathode salt. (AgF, AgCN, AgBr, AgNO3, AgI, or AgClO4)
4e. After 90 sec insert the cathode contact and connect multimeter.
4f. After 130 sec start data logging. Logging program will record current output of the cell every 4 seconds over a duration of 3600 seconds.
4g. Repeat 4 times with each cathode salt.

(5) Effect of water concentration on gradient formation.
5a. Insert a cleaned 1 cm piece of lithium into the syringe barrel and compact the wire so it forms a tight seal at the edges of the syringe and is in electrical contact with the copper anode contact.
5b. Add 0.3 ml of the 0.2 g/ml electrolyte solution with variable water content (0.2%, 0.1%, or 0.01%).
5c. Inject one separator into the lithium cell. Subject cell to light vacuum to remove argon bubbles from separator.
5d. Start timer and add 0.2 ml AgNO3
5e. After 90 sec insert the cathode contact and connect multimeter.
5f. After 130 sec start data logging. Logging program will record current output of the cell every 5 seconds over a duration of 3600 seconds.
5g. Repeat 4 times with each electrolyte of variable water concentration.

Data: see archive

http://www.rtftechnologies.org/Design/Assets/files/licell/

Results

• Fig 4-1 shows the average output current of a Li/AgNO3 cell as electrolyte concentration is increased. Output current increases up to a point as electrical resistance decreases, however it begins to decrease as the electrolyte viscosity begins to impede ion travel resulting in concentration polarization of the electrolyte.

• Fig 4-2 shows an increase in electrolyte conductivity as the electrolyte concentration increases.

• Fig 4-3 shows an increase in electrolyte viscosity as electrolyte concentration increases. This increase in viscosity impedes ion travel increasing the internal resistance of the cell.

• Fig 4-4 shows an increase in peak output current as anion radius increases. This increase in current reflects the mobility of the anion in the cell’s electrolyte. All anions used in this experiment have a –1 charge, only their radii change. As an anion’s radii increases, it’s volume increases as well, decreasing its charge density. Since water has a dipole moment, it is attracted to a given anion with a force directly proportional to the anion’s charge density. Smaller anions have a greater charge density and attract more water molecules increasing its effective volume and therefore decreasing it’s diffusion rate.

• Fig 4-5 shows that as water contamination in the electrolyte increases, peak current output current decreases. This effect is due to the water’s attraction to the anion, increasing the effective volume. This effective increase in the anion’s volume reduces its diffusion rate, increasing the magnitude of concentration polarization in the electrolyte.

• Fig 4-6 shows the averages of the peak currents and steady state currents recorded during 5 trials for each anion. As anion radii increased, both the averages of peak and steady state current increased due to the lower magnitude of ion gradients of the larger anions.

Discussion

In this research project five sets of experiments were conducted to analyze ion gradient formation in Li/Ag-salt cells. These were:
1. Effect of electrolyte concentration on current output;
2. Effect of electrolyte concentration on electrolyte viscosity;
3. Effect of electrolyte concentration on electrolyte conductivity.
4. Effect of ion radius on ion gradient formation and steady state current.
5. Effect of water on ion gradient formation and steady state current

1. Effect of electrolyte concentration on current output
The results of set (1) experiments show average current output over a 400 sec duration as electrolyte concentration increased from 0.0 g/ml to 0.4 g/ml in 0.1 g/ml increments. As electrolyte concentration increased, current of Li-AgNO3 cells, used as standards, increased from 0.012 mA at 0.0 g/ml to a peak of 1.585 mA at 0.2 g/ml and then decreased to 0.15 mA at 0.4 g/ml. This peak occurred at 0.2 g/ml due to optimum ion transference rates at this concentration. The decrease in average current at electrolyte concentrations greater then 0.2 g/ml is caused by the increased rate of concentration gradient formation in the electrolyte caused by decreasing transference rates due to the increase in electrolyte viscosity, (Fig 4-3). The increasing viscosity of the electrolyte decreased ion mobility and the rate of ion diffusion of the cell causing areas of higher ion density localized at the surfaces of the anode and cathode.

2. and 3. Effect of electrolyte concentration on electrolyte viscosity and electrolyte conductivity
The results of set (2) and set (3) experiments show electrolyte viscosity and electrolyte conductivity, respectively, as electrolyte concentration increased from 0.0 g/ml to 0.4 g/ml in 0.1 g/ml increments. As electrolyte concentration increased, conductivity and viscosity increased in an exponential manner. In the lithium cells tested, (Fig 4-1), the effect caused by the increase in conductivity was canceled out by the increase in viscosity at 0.2 g/ml in regards to the average current output over a 400 sec interval. This is due to higher viscosity electrolytes yielding lower rates of ion diffusion. This causes an increase in the formation rate of concentration gradients. These gradients reduce the current output of the cell.

4. Effect of ion radius on ion gradient formation and steady-state current
In this set of experiments, the rate of gradient formation increased as ion radius increased. Since the electrolytes used were only 99.8% anhydrous, as tested on a Karl-Fisher titratior, the relatively large amount of water in the electrolyte impeded the motion of small ions, causing the formation of gradients of a high magnitude near the cathode resulting in lower initial current.

As ion radius decreased (Fig 5-1), charge density of the ion increased since the anionic charge was localized in a smaller volume. Since water has a dipole moment, it is attracted to the anions with a force proportional to the charge density. Smaller ions with higher charge density attract more water molecules increasing the effective volume of the ion and therefore have lower transference rates than larger ions with lower charge densities. This effect can be seen in the results illustrated in Fig 4-4. As ion radius increased, most cells show a higher peak current and a faster decrease in current until current approached a steady state. At this point the cell reached equilibrium, when the ion diffusion rate became equal to the rate of ion production.
The peak current was caused by the increase in dissolved cathode salt as the cell was run. The dissolved cathode salts would be in contact with the cathode electrode, therefore increasing the effective surface area possible to accept electrons from. This increases the current output until ion gradients of an effective magnitude form, therefore limiting current. The formation rate of these gradients would depend on the amount of water in the electrolyte as well as the current flowing through the cell.
It should also be noted that as the cell runs, the electrolyte slightly increases in concentration. Although the results in Fig 4-1 suggest that peak current occurred at 0.2 g/ml LiBF4 in a solution 1:1 by volume of 1,2-dimethoxyethane and propylene carbonate, it should be noted that the current was determined at electrolytes between 0.0 g/ml and 0.4 g/ml in 0.1 g/ml steps. It is known from the data in Fig 4-2 that when electrolyte concentration increased, conductivity increased as well. Therefore it is reasonable to assume that the conductivity of a given electrolyte will increase as the cell is run, increasing current output, before ion gradients cause a noticeable increase in electrolyte resistance and the cell starts to approach steady-state current.

5. Effect of water on ion gradient formation and steady-state current
In this set of experiments, the gradient formation increased as the water in the electrolyte increased as seen in fig 4-5. It was observed that as water concentration increased, peak current decreased. The increase in water concentration causes ion gradients to increase demonstrating that water contamination in the electrolyte will reduce output current due to the attraction between anion and water.

In these experiments lithium cells were constructed to test the following hypotheses:

• The ion migration rates in Li | Ag-salt cells will increase as anion radii increases.
• Li | Ag-salt cells with a given anion will have decreasing ion transference rates in a LiBF4 electrolyte of increasing concentration.

Increasing the LiBF4 concentration in the electrolyte of a Li|AgNO3 cell increases the formation rate of gradients. This causes a resistance increase in the electrolyte and therefore a decrease in current output. The increasing viscosity of the electrolyte decreased ion mobility and the rate of ion diffusion in the cell causing areas of high ion concentration localized at the surfaces of the anode and cathode.
Ions with small radii showed a high initial rate of gradient formation. This effect is due to the water in the electrolyte impeding the motion of small ions, causing the formation of gradients of a high magnitude near the cathode. The presence of these gradients resulted in a lower initial current (Io) delaying peak output due to the slow rate at which ion pairs were produced. Ions with larger radii show a higher current peak due to the higher initial transference rate.
Further research of these properties in variably hydrated electrolytes have shown that an increase in water in the electrolyte causes the current output to decrease since the water is attracted to the anion. This increase in the anion’s effective volume causes a decrease in ion transport away from the cathode, increasing the rate at which ion gradients form.
The results of this research showed that water contamination decreases the current output of cells utilizing ions with smaller radii and that electrolyte concentration in addition to ion radii effects the magnitude of ion gradients in the electrolyte. The optimum Li|Ag-salt cell would have a 0.2 g LiBF4 / ml electrolyte, an anion with the maximum practical radii and a minimum amount of water in the electrolyte.

This research was conducted during the months July-October 2002. Laboratory space was provided by RTI International. Supplies and assistance were provided by Herbert Seltzman. Professor Peter S Fedkiw and Ruchi Singhal of the Chemical Engineering Department, NCSU, provided background information on gradient formation in lithium cells as well as guidance on formatting a research paper.

[1] Bruce, Peter G., James Evans, and Colin A. Vincent. (1988). Conductivity and Transference Number Measurements on Polymer Electrolytes. Solid State Ionics: 918-922.

[2] Jenkins, Donald B., Jack Passmore, Leslie Glasser, and Helen K.
Roobottom et al (1999). Thermochemical Radii of Complex Ions. Journal of Chemical Education. N.p.: The Division of Chemical Education of the American Chemical Society, 1570-1573.

[3] Mayes, Anne M., and Donald R. Sadoway. (2002). Portable Power: Advanced Rechargeable Lithium Batteries. MRS Bulletin: 590-592. 3 September 2002. <http://www.mrs.org/publications/bulletin>.

[4] Linden, David. Handbook of Batteries. (1995).2nd ed. Washington, D.C.:
McGraw-Hill, INC, 2.15-2.18.

[5] Crompton, TR. (1995). Battery Reference Book. . N.p.: Butterworth-Heinemann Ltd,

[6] Handbook of Chemistry and Physics. (1976) . N.p.: CRC Press.
Handbook of Chemistry and Physics. (1959). N.p.: Chemical Rubber Co.

[7] Heise, George W. (1971). The Primary Battery. 1st. Ed. N.p.: John Wiley and Sons P.

[8] "Lithium." (1993). Compton's Interactive Encyclopedia. N.p.: Compton's NewMedia, Inc. N.

[9] Fedkiw, Peter S., (April - November 2002). Personal communications. Department of Chemical Engineering, NCSU, Raleigh, NC.

By attempting to reproduce any experiments or devices listed on this domain in part or in whole, you agree to hold me harmless against any lawsuit or liability.

Copyright © 1998 - 2005 by Andrew Seltzman. All rights reserved.