FN Tandem Van de Graaff Accelerator

The model FN Tandem Van de Graaff accelerator of the Nuclear Science Laboratory at the University of Notre Dame has been in service since its installation in 1968. Originally, it was capable of producing accelerated positive ion beams and accelerated electron beams, but it was modified in the middle 1970s to remove the electron beam capabilities, and has been used exclusively for positive ion beam acceleration since that time.

The fundamental principles of the FN Tandem accelerator are straightforward. A centralized metal electrode, known as the terminal or spinning, is charged to a very high positive potential. A negatively charged ion beam, produced by ion sources external to the accelerator, is transported in vacuum toward the terminal, accelerating to high energy as it approaches the terminal. As the beam enters the region which houses the terminal, it passes through a thin carbon foil which strips electrons from the ions in the beam, leaving the beam positively charged. This positively charged beam accelerates away from the high positive potential at the terminal, and exits the accelerator at very high energy. The name "Tandem" arises from the two accelerations (one before stripping and one after) that the ion beam experiences.

The FN Tandem accelerator is housed within a large (40 ft long, 12 ft diameter ) steel tank, seen in the photograph above, that serves to isolate the high voltage surfaces of the accelerator electrodes from the outside world. Additionally, the tank is filled to high pressure with a special insulating gas, which helps to prevent electrical discharging of the high voltage surfaces within the accelerator.

The structures within the tank which support the terminal electrode are referred to as "columns". The low energy (LE) column extends from the tank base nearest the ion source to the center of the terminal region, and the high energy (HE) column extends from the terminal region to the tank base opposite the ion sources. These columns are composed of a series of metal planes, with each plane electrically isolated from the next by 4 glass blocks, which are glued to each plane. There are approximately 200 such planes in each of the LE and HE columns. A very large spring inside the HE tank base compresses both columns with tremendous force, holding the columns and the terminal region suspended within the tank.

The evacuated acceleration tubes through which the beam passes are mounted along one side of the column structure, as can be seen in the photograph at left, which shows a small section of the column and one of the acceleration tubes, viewed from the side. Some hardware has been removed to make the column and tube visible. There are four separate acceleration tubes within the FN Tandem, each measuring approximately 8 ft long and 8 inches in diameter, mounted end to end with metal bellows connections. The tubes are constructed of 1 inch long sections glued together, with each section being comprised of a hollow cylindrical glass section glued to a "dish shaped" metal electrode. These electrodes contain central apertures through which the beam passes. The tubes are mounted along the columns so that the electrodes in the tubes are directly opposite the metal planes in the columns. The acceleration tubes are mounted so that they are electrically isolated from the columns, but each electrode in each tube is electrically connected to the corresponding plane in the column. The dark color along the acceleration tube is due to the radiation the tube is exposed to while the beam is being accelerated. Also notice the many "acorn nut" fixtures attached to each of the metal electrodes in the accelerating tubes. These are used to form spark gaps, so that if a sudden electrical discharge occurs, the sparks will form across these spark gaps, rather than through the glass in the column or the tube, which otherwise could be seriously damaged.

In the fall of 2000, the system for charging the terminal of the FN Tandem accelerator was upgraded from the original belt charging system to a "Pelletron chain" charging system, supplied by the National Electrostatics Company, Inc. (NEC). This robust charging system is in use in many currently operating accelerators, and allows for a very reliable and stable charging of the terminal electrode. The system consists of a chain formed by tubular metal links connected to each other with nylon inter-connectors, so that each link in the chain is electrically isolated from the next link. The chain is mounted along the side of the column, and extends from the tank base to the terminal. In the FN Tandem accelerator, there are two completely separate chain systems, one along the LE column and one along the HE column.

The chain travels through the accelerator at a speed of approximately 40 mph, driven by a motor and pulley arrangement at the tank base end. The links are charged by induction as they pass near special metal electrodes maintained at high voltage (variable, up to 60 kV). As the chain moves, each positively charged link travels to the terminal where the charge on each link travels through the chain pulley to the terminal electrode. As the chain leaves the terminal region to travel back to the tank base, it is again charged by induction, although the charge is now negative. This allows both the up and down runs of the chain to contribute to the charging of the terminal. For a more detailed explanation of the principles of the pelletron, please see the excellent description by NEC at

To insure uniform acceleration of the ion beam as it passes through the FN Tandem, it is necessary to control the electric potential within the accelerator, which varies from ground potential at the LE and HE tank bases to as much as 10 MV at the terminal. A uniform gradient is required, and this is accomplished by connecting resistors between each plane in both columns, so that each column acts as a voltage divider circuit to ground. In our case, a pair of 300 Megohm resistors are connected in series between each plane, so that the total resistance across the gap between each plane is 600 Megohms. In this photograph, which is a view of the column from above, the resistor mounts are visible. Each resistor is a small ceramic core with a thin film coating, about the size of a common pencil. Each resistor is housed in an aluminum tube, to shield the resistors. Since the resistance across each gap in the column is the same, the voltage varies linearly from ground to terminal voltage across the length of each column, producing a constant gradient suitable for accelerating the ion beam. Therefore, when the terminal is at voltage, current flowing through the column resistors biases each of the column planes, which in turn bias each electrode in the acceleration tubes, so that the constant gradient field is present within the acceleration tubes.

In the photograph, students are completing the re-assembly of the column structure by replacing the tubular stainless steel hoops which surround each column plane, also encircling the acceleration tubes, and act to preserve the equipotential nature of the field at each column plane. The temporary lights above and the floorboards below are removed when the accelerator is in operation.

Successful operation of the accelerator for experiments in nuclear physics requires that the terminal voltage remain very constant and stable over long periods of time. Equilibrium must be established between the charge brought to the terminal by the pelletron chains and that which flows from the terminal to ground through the column resistors. Also, small irregularities in the links of the pelletron chain will tend to cause variations in the charge delivered to the terminal, and this must be compensated for. This is done through the use of corona points, a collection of a dozen or so very sharp metal needles attached to the end of a moveable arm. The arm containing the corona point assembly is mounted to the tank wall directly opposite the terminal, so that the points can be extended toward or extracted away from the terminal. During accelerator operation, the corona points are moved close enough to the terminal so that a coronal discharge begins at the points, and this discharge causes charge to flow from the terminal through the corona points. A variable resistor within the electrical circuitry connected to the corona points is adjusted to increase or decrease the charge extracted from the terminal so that a constant terminal voltage is maintained.

The terminal voltage is continuously measured in real time with a device known as a generating volt meter (GVM), which is mounted in the tank wall directly opposite the terminal. The GVM has a set of stationary metal vanes mounted behind a set of rotating metal vanes. When the accelerator is operating, the GVM is exposed to the electric field at the location of the terminal, and the capacitance of the GVM varies as the rotating vanes alternately cover and then expose the stationary metal vanes to the electric field. This measurement from the GVM can be then be used to determine the terminal voltage. Two devices known as capacitive pick-off (CPO) units are mounted 180 apart in the tank wall near the terminal. These devices measure any variation in the capacitance due to motion of the terminal, and the signal from these CPO units is used to correct the GVM signal for any terminal motion (FN charging schematic).

Maintaining a constant, stable terminal voltage is absolutely essentially for nearly every experiment performed with the FN Tandem accelerator. As discussed above, the variable resistance in the corona points assembly is continuously adjusted to control the corona current so that a constant terminal voltage is maintained.

The signal used to adjust the variable resistance in the corona points assembly is provided by the stabilizer circuit, and is generated in one of two ways. When the stabilizer circuit is set to GVM Mode, the output of the GVM is compared to a reference which is set by the experimenter to the desired terminal voltage. The error signal created from the difference between the reference and GVM signals is then used to adjust the variable resistance in the corona points assembly, which causes the terminal voltage to change until the reference and the GVM signals agree.

When the stabilizer is set to Slit Mode, an error signal is generated by a set of slits located at the exit of the 90 analyzing magnet, which is downstream from the accelerator. The slits are set symmetrically within the beam line, with one slit on either side of the beam. The magnetic field in the analyzing magnet is adjusted to allow only that beam with the energy selected by the experimenter to complete the 90 bend. When the magnetic field in the analyzing magnet is correctly adjusted, the beam will pass through the center of the gap between the slits. The slits are set to intercept a small amount of beam from the outer edges of the beam envelope, and so a well centered beam will strike both slits equally, so that the same amount of charge is deposited on each slit. However, if the beam energy varies slightly due to variations in the terminal voltage, then the beam will not have the correct energy to traverse the 90 bend, and more beam will strike one of the analyzing slits than the other. An error signal is generated based on the difference in the slit current readings, and this signal is then used to adjust the variable resistance in the corona points assembly.

The FN Tandem accelerator requires external ion sources to produce negatively charged ion beams. These ion beams are typically singly charged, with each ion containing one extra electron. These negatively charged beams are produced by ion sources which are maintained at large negative potentials, so that the ion beam energy as it enters the FN Tandem accelerator is typically 40 keV for helium beams and 80 keV for other ion beams. The beam accelerates through the evacuated acceleration tubes toward the central terminal electrode in the FN Tandem, which is maintained at a large positive potential determined by the experimenter. The energy of the beam at the terminal is then approximately T MeV, where T is the terminal voltage in MV. As the ion beam reaches the terminal, it enters a region known as the "stripper", where a very thin (approximately 3 micrograms/cm2 ) carbon foil is housed in the acceleration tube assembly. The ion beam passes through this stripper foil, and interactions between the atoms in the foil and the beam strip electrons from the ions in the beam. The number of electrons stripped from the ions in the beam varies with the beam energy, and there is usually some distribution of positive charge states present in the ion beam exiting the stripper region. The resulting ion beam, which is now positively charged, accelerates away from the large positive potential at the terminal electrode toward the far end of the accelerator. The final energy of the beam is then T + QT MeV, where again T is the terminal voltage in MV and Q is the positive charge state of the beam exiting the stripper region. These two accelerations, that of the negatively charged beam toward the terminal and that of the positively charged beam away from the terminal, give rise to the name Tandem.

For example, consider the acceleration of an oxygen beam when the terminal voltage is 10 MV. The beam entering the accelerator is a singly charged negative oxygen beam, so that the beam energy at the stripper is 10.080 MeV (10 MeV from the acceleration and .080 MeV from the injection energy provided by the ion source). At the stripper, positive oxygen ions will be produced ranging in charge state from 1+ (two electrons removed from the original beam) to 8+, with a peak in the distribution at about 5+. The 5+ beam will gain 50 MeV (5e .10 MV) as it accelerates toward the far end of the accelerator, and will exit the accelerator with an energy of 10.080 + 50 = 60.080 MeV. By adjusting the terminal voltage and selecting the appropriate charge state exiting the stripper, oxygen beams of any energy from approximately 2 MeV to approximately 100 MeV can be produced.

In 1991, a second carbon foil stripper region was added at the midpoint of the high energy column of the FN Tandem accelerator. Since the ion beam has accelerated from the terminal to the midpoint of the high energy column, the energy of the beam at this location is significantly higher than the beam energy at the terminal stripper. Stripping at these higher energies increases the population of the more heavily stripped charge states, resulting in more intense beams at higher beam energies. Consider our previous example of an oxygen beam accelerated with 10 MV on the terminal. The beam energy at the terminal is 10.080 MeV, as before, and most of the beam will exit the terminal in charge state 5+. At the midpoint of the high energy column, the beam energy will be 10.080 MeV + (5e . 10 MV) = 35.080 MeV. If the second stripper is used, then the beam will strike the second stripper foil with an energy of 35.080 MeV, which is much higher than the energy of the beam at the terminal stripper. At this energy, most of the beam will be fully stripped to charge state 8+. This fully stripped beam accelerates down the remainder of the high energy column, exiting the accelerator with an energy of 35.080 MeV + (8e . 10 MV)= 75.080 MeV. This to be compared with the previous result of 60.080 MeV without the use of the second stripper. To date, this second stripper has been used to produce several high energy, highly stripped ion beams, including 110 MeV chlorine in charge state 14+.

<< Back to Research Facilities

Site Map | Search | Contact Us | Department of Physics | University of Notre Dame