The History

The demonstration was devised by one of us (PK) in 1986, after reading an activity in Hubert Alyea's 1947 compendium of chemical demonstrations from this Journal (1). In that activity, Alyea hooked a magnesium strip to the negative battery terminal of an electric bell and hooked a copper strip to the positive terminal. He placed the loose ends of the strips into a 1M ₂SO₄ solution and the bell rang. After trying the demonstration, it seemed to make sense to modify the electrolyte to orange juice because it is safe, readily available, and would be a mixture in which the magnesium would oxidize more slowly than in sulfuric acid. Further, a clock was substituted for the bell because a clock is easier on the ears than a bell. A video of the orange-juice clock setup is given as Figure 1.

Figure 1.The orange juice clock set up. Video of orange juice clock was filmed and editted by Jerry Jacobson at the University of Wisconsin - Madison. (larger version, 825 kB)

The apparatus was presented in 1987 as part of a teacher workshop led by Irwin Talesnick, then of Queen's University in Canada. Talesnick, whose distinguished career has been characterized by seeing educational possibilities in so many things, created a modified version of the clock, with the atomic numbers of the elements representing the hours in the day (see Fig. 2) in his internationally popular workshops. Due largely to Talesnick's efforts, the orange juice clock is a standard demonstration in many chemistry programs and presentations.

Figure 2.Irwin Talesnick represents the hours of the day by the corresponding elements in his clock.

The Procedure

This can be done as a demonstration or as an activity, although at about $10 per clock, expense does become an issue. There are no unusual safety precautions with this demonstration. We know of no accidents that have occurred with the orange juice clock. The demonstration requires:

a single AA-cell battery-operated wall clock with a sweep-second hand

a medium-sized beaker (600 mL is fine)

enough orange juice or other electrolyte mixture or solution to fill the beaker about 2/3 full (tap water often works fine!)

a 20-30-cm magnesium strip, coiled at one end or wrapped around a popsicle stick

a 20-30-cm copper strip, coiled at one end

alligator clips to connect the strips to the battery terminals on the clock

a stand against which to lean the setup

The demonstration is put together as shown in Figure 3. Connect the magnesium to the "-" contact of the clock and the copper to the "+" contact. Immerse the other ends of the strips into the solution. The clock will start to tick within a few seconds. If it does not work within a short period of time, check that the strips are well connected to the battery terminals, are hooked to the proper poles, and are not touching each other. The clock should keep reasonably close time (in orange juice) for a couple of days, or until the magnesium is nearly completely oxidized.

Figure 3.A schematic of the orange juice clock seup.

Video of orange juice clock. In video, the copper electrode is on the left and the magnesium electrode is on the right. Video was filmed and editted by Jerry Jacobson at the University of Wisconsin - Madison. (larger version, 858 kB)

The Chemistry Basics

When we ask students or precollege teacher groups about the reduction and oxidation reactions that are occurring, they invariably answer that the magnesium metal is being oxidized and the copper metal is being reduced. This response is important because we use it to impress upon students and workshop participants the importance of looking carefully at the system before giving what might seem like an obvious answer. The copper cannot be reduced because there is no copper ion in solution, and transition metals cannot be reduced to anions. Given what is actually in solution, participants can conclude that hydrogen ion can be reduced to molecular hydrogen (in orange juice) or that hydrogen in the water molecule is being reduced to molecular hydrogen (in hard tap water). In distilled water, the clock does not run because the internal resistance of the solution is too high, thus forcing the current to be very small. The reactions of interest are given as eqs 1-3:

oxidation: Mg -> Mg²⁺ + 2e^- E^o = 2.37 vs. SHE (1)

reduction 2H⁺ + 2e^- -> H₂ E^o = 0.00 vs. SHE (2)

(acid solution)

reduction (water) 2H₂O + 2e^- -> H₂ + 2OH^- E^o = -0.8277 vs. SHE (3)

where E^o = the voltage under standard conditions and SHE = standard hydrogen electrode.

At standard conditions, under zero load (all activities equal to one and 298 K) the cell voltage should theoretically be 2.37 V in acid (pH = 1) and about 1.54 V in neutral solution, either of which is enough to allow the clock to run. It is important to remember the IUPAC convention for electrochemical cells: that voltage of the cell equals voltage of the cathodic half-cell minus voltage of the anodic half-cell. In this case,

E^o = 0.00 V - (-2.37 V) = 2.37 V

The standard free energy calculation is straightforward in each case (eq 4),

DeltaG^o = -nFE^o

in which n = number of moles of electrons transferred, as dictated by the stoichiometry of the reaction (in all reactions above, n = 2); F = Faraday's constant, 96,498 C per mole of electrons (it is useful to show students that this number is equal to the product of Avogadro's number and electron charge); and E^o = cell voltage under standard conditions (E^o = 2.37 V = 2.37 J/C in acid solution of pH = 1). In acid solution, DeltaG = -457,000 J = -4.57 kJ. The reaction is spontaneous (and there is enough current flow). The clock ticks.

This represents an overview of the fundamental chemistry, suitable for a workshop, high school, or non-science first-year college audience. The discussion below considers some more advanced aspects of the demonstration, which make this an excellent demonstration for the first-year science majors' course as well as upper- and graduate-level analytical and electrochemistry courses.

For Those Who Want More

More advanced students can readily explore the parameters of the clock system beyond merely studying cell voltage at standard conditions. In this system, for example, two sources contribute to the oxidation of the magnesium electrode. One is the reaction with acid as part of the process that runs the clock. Also present is the reaction in acid solution that occurs irrespective of the electron flow used to run the clock, a process of corrosion that dissolves the metal without useful energy being obtained. It is possible to distinguish between the two and to determine, via Faraday's constant, the average current available to the clock in this system.

Faraday's Constant and the Average Current

Data for a typical determination are given in Table 1. The data were taken using a 0.3317-g magnesium strip that had been cleaned with steel wool. The magnesium and copper strips (the copper was cleaned by dipping in 1 M nitric acid for a few seconds) were placed in 400 mL of a commercial orange juice so that about 15 cm of each strip was above and 15 cm below the liquid line. About 5 cm of the magnesium strip was coiled. The clock was hooked up in the usual fashion. At 1-hour intervals, the mass of the magnesium wire was determined on an analytical balance. A digital multimeter was used to measure the voltage every hour and the current every other hour. The pH of the juice, initially 3.85 at 20 °C, was 3.93 at the end of the experiment, as measured with a portable pH meter. This is a typical result. A comparison system (called "no clock" in Table 1) was set up merely by putting a 0.3317-g strip of magnesium in 400 mL of orange juice. With this system we can exemplify a "corrosion process" where the anode and the cathode are in the same place. There is consumption of magnesium and evolution of molecular hydrogen but no useful current can be obtained. The mass of this magnesium strip was measured at 1-hour intervals.

Faraday's constant, which relates coulombs to moles of electrons, can be used to calculate the approximate current available to the clock in this system. The current will not be constant because the H⁺ concentration (related to pH) is changing and also because the surface area and composition of the magnesium electrode change with time. The change is not necessarily regular, because although the surface is being oxidized, it is not smooth. The actual available surface area will therefore be considerably greater than the geometric surface. The mass of magnesium oxidized in the clock reaction over the 4-hour period is approximately equal to the change in grams of magnesium while running the clock minus the mass of magnesium oxidized in orange juice without the clock. Using the data from Table 1,

grams Mg oxidized to run clock = approx. (0.3317 - 0.3089)

- (0.3317 - 0.3136) = 0.0047 g Mg

The average current can then be calculated via Faraday's constant:

This is a rather simplistic way to get the current, but it shows well the use of Faraday's constant.

The Value of Computer Interfacing - Exploring the Physics of Current/Voltage Measurements

A more instructive measure of voltage vs. time, which opens up the activity to more interesting possibilities, was obtained by interfacing the clock to a Macintosh 8100/80 microcomputer via Vernier Corporation serial box interface hardware and software (see ref 2). This affordable ($100-$250 per computer) interfacing package is being used in our first-year chemistry laboratories. The interfacing setup permitted data to be acquired at the much more meaningful rate of up to 50 points per second. It also permitted us to observe voltage variations with time while the strips in orange juice were hooked up to the clock. When data are taken 50 times per second rather than once every hour, the data take on new meaning. Figure 4 shows that there is a substantial drop in voltage each time the clock ticks.

Figure 4.The observed voltage drop in the circuit corresponds to the ticking of the clock. The drop is due to the internal resistance of the orange juice solution.

Video of orange juice clock connected to chart recorder demostrating voltage drop in the circuit as clock ticks. Video was filmed and editted by Jerry Jacobson at the University of Wisconsin - Madison. (larger version, 1.2 MB)

This observation can be explained and can be predicted as part of a student activity, if we understand the nature of an open vs. a short circuit.

A battery can, in concept, perform between two extreme points: an open circuit, in which the voltage (V) is at a maximum but there is no current (I), and a short circuit, in which the current is at a maximum but there is no voltage. A battery is best used at an intermediate point where the power, I xV, is a maximum. In summary (eqs 5-7):

(5) Open Circuit: V = maximum and I = 0

(6) Short Circuit: I = maximum and V = 0

(7) Battery Use: I x V = power = maximum

A 1.5-V battery has an open circuit potential of 1.5 volts. When the battery is working, however, the real voltage will be less than 1.5 V. This is due to the internal resistance of the battery. So the real voltage of the battery (V_real) equals the open circuit voltage (V_open) minus the voltage drop due to internal resistance in the battery. This drop is equal to the current passing through the circuit (I) multiplied by the internal resistance of the battery (R_int), as shown in eq 8:

V_real = V_open - I x R_int (8)

If the current passing is 0.002 A and the internal resistance of a 1.5-V battery is 50 Ohms, the real voltage is 1.4 V:

V_real = 1.5 V - (0.0020 A x 50 Ohms) = 1.4 V

In this activity, in which we make a battery with a magnesium and a copper strip in orange juice, the juice itself provides the internal resistance in the battery. The key then to determining what the voltage drop should be is to find the internal resistance of the orange juice and then to find the current passing through the clock circuit.

Students can determine the internal resistance of the orange juice by performing the following measurements. Note that the internal resistance of the orange juice is highly dependent upon how far apart the strips are in solution. The strips should be firmly taped, top and bottom, to the beaker. The data below were typical for 0.35-g Mg and 6.0-g Cu strips that were 4 cm apart in a 600-mL beaker with 400 mL of orange juice. The solution was not stirred. The area of the strips in solution was about 7.5 cm² for the magnesium and about 15 cm² for the copper.

1. Measure, using a high-impedance voltmeter, the voltage of the circuit using the voltmeter itself, rather than the clock, to complete the circuit. This will give a good approximation of the open circuit potential (the current is negligible, on the order of microamps if the voltmeter has MOhm resistance). In our setup Vopen = 1.772 V.

2. Attach a 1000-Ohm resistor across the circuit. Measure the voltage in parallel to the resistor. This voltage (1.037 V in our setup) will be equal to the current in the circuit x the resistance of 1000 Ohm. We can now solve for the current in this circuit:

I = V/R = 1.037 V/1000 Ohm = 0.001037 A = 1.037 mA

3. The resistance of the orange juice is then calculated via the difference between the open circuit voltage (1.772 V) and the voltage with a known resistance (1.037 V). The difference, 1.772 - 1.037 = 0.735 V, equals the product of the circuit current and the resistance of the orange juice (ROJ), or

R_OJ = 0.735 V/0.001037 A = 708 Ohm

4. Finally, measure the current that the clock itself requires by hooking up in series an ammeter to the battery and the clock. The reading is not easy to take with an ammeter, which does not sample very often, and integrates across time. The computer interface works better for this. In our clock, a current of 0.49 mA was used.

5. The payoff comes at this point. The predicted voltage drop (I x R_int) can be calculated,

voltage drop = I x R_int = 0.00049 A x 708 Ohm = 0.35 V

Our observed voltage drops for this system were typically around 0.30 V. As a confirmation of the relationship of internal resistance to voltage drop, we placed the strips 1 mm apart in an orange by digging 2 holes in the orange and placing into the holes the coiled parts of the strips. We expected the voltage drop to be much higher than with the juice, due to the much higher internal resistance of the orange. Even when the strips were nearly touching, the drop was about 1 V.

The Water Clock

We discussed above the difference in the redox system when water is used rather than orange juice. Distilled water, which has a high internal resistance, will not permit the clock to run. However, hard tap water or distilled water with, for example, 1 g of table salt in 300 mL of water will work fine. As expected, because of the lower hydrogen ion concentration, the initial cell voltage is lower, typically around 1.45 V. The clock also ticks more slowly and more softly in water than in orange juice. In water, a black precipitate forms on the magnesium electrode and becomes more extensive with time. When the strip is removed from distilled water, and allowed to dry the precipitate turns white. Further student exploration on the precipitate might include designing experiments to find out if the precipitate is a carbonate or an oxide (from the hydroxide.)

Non-Nernstian Considerations

The systems above were always run without stirring because when setting up demonstrations, portability, simplicity, and expense are important, and the main concepts are as clear with a stir bar as without. We do note, however, that when the solution is constantly stirred, the rate of magnesium oxidation both with the clock setup and simply in solution is considerably faster than when the process is diffusion-limited. In fact, whereas the Mg strip will often last for several days in very dilute acid and overnight in orange juice when the solutions are not stirred, it will break off within 4 h when the solutions are stirred.

Another important issue relates to our use of the Nernst equation to account for the potential developed in the system. This equation is very useful to assess chemistry at equilibrium conditions, but the orange juice clock is using an electric current and so is not at equilibrium. In our orange juice system, the Nernst equation (eq 9) is,

in which E° = E°_Mg - E°_H2 - h

and h = overpotential = difference in H₂/H⁺ couple at a copper electrode minus that at platinum black. The hydrogen overpotential on a copper surface is typically 0.23 V. Another treatment of overpotential is given below.

In our experiments, magnesium concentration and hydrogen activity were not measured or controlled; the pH was 3.85-3.93, as described above.

The maximum theoretical potential of the electrode system is greater than that which is available to the clock when there is current flow. As described above, the potential drop is calculated as current ¥ internal resistance, equaling the "iR drop." This is why potentials are measured with a voltmeter, with a very high internal resistance, which draws very little current from the system.

The measure of how far a system is from equilibrium is called the overpotential (h)

h = actual potential minus potential at equilibrium

This, along with the anodic and the cathodic components, the energy involved, and the temperature of the system are all dealt with using the Butler-Volmer equation, given as eq 10:

I = i_o(e^+Fh/2RT - e^-Fh/2RT) (10)

in which i_o is a specific constant for every system "electrode-electrolyte" and is called the "equilibrium exchange current." For this case, we have assumed the symmetry factor to be equal to 1/2. A detailed discussion of this factor is beyond the purpose of this paper, but can be found in ref 3.

Questions To Raise with Students/Teacher Workshop Participants

This demonstration can be a starting point for many concepts. It is especially powerful in showing how chemistry can be used beyond the chemistry laboratory. The primary question is "how is this system different from that in which redox occurs at one surface (such as a zinc strip placed in a solution of copper sulfate)?" The key with this electrochemical cell is that we are separating the anode from the cathode to take advantage of the electron flow (current) through an external wire and this current will give power to the clock or any other device. This is the essence of a battery. Other questions we often ask are:

Is enough current produced to run a small electric motor? Light a light?

Would the system work if we put Cu²⁺ ions into the solution?

What would happen if we titrate the acid solution with a strong base while the clock is running?

What happens to the voltage if we put Mg²⁺ into the system?

Are the complex ions of Mg²⁺ with citric acid important to the potential value?

What is the relationship between the clock ticking and different juices? Why is this relationship so?

What are the reactions in the most popular commercial batteries?

What is the chemical basis of rechargeable batteries?

Related Activities

We do this demonstration while studying electrochemistry during the second semester of the general chemistry sequence. In precollege teacher workshops, it is an important focus of an Operation Chemistry unit dealing with energy needs for living on board the space shuttle (4). A number of fairly safe activities work well as lead-in or follow-up material, as described in Table 2. Copies of these activities are available by writing to PK.

Acknowledgments

We wish to thank Walt Hancock and Jonathan Skean, along with our wonderful undergraduates Mickey Richards, Cory Emal, Julie Grundman, Jeff Atkins, and Darren Jack, for being there.

Literature Cited

1. Alyea, H. N. Tested Demonstrations in General Chemistry, 1955-1956; American Chemical Society: Washington, DC, 1956.

2. Vernier Software, 8565 SW Beaverton Hillsdale Highway, Portland, OR 97225; phone (503) 297-5317.

3. Bockris, J.; Reddy, A. K. N. Modern Electrochemistry; Plenum: New York, 1970; Vol 2.

4. Kelter, P.; Hughes, K.; Murphy, A.; Roskos, P. J. Sci. Teacher Educ. 1995, 6, 57-59.

5. Tested Demonstrations in Chemistry; Gilbert, G., Ed.; American Chemical Society: Washington, DC, 1994; Vol. 1, #E-13.

6. Katz, D. A.; Willis, C. J. Chem. Educ. 1994, 71, 330-331.

7. Holmquist, D. D.; Volz, D. L. Chemistry with Computers;Vernier Software, Portland OR, 1994.