SPMs Analyze Disk Surfaces

Scanning-probe microscopes reveal surface and magnetic characteristics of magnetic media.

Jon Titus, Editorial Director -- Test & Measurement World, 8/1/1999

AFMs and MFMs play a role in rooting out problems that defy careful examination by optical methods. Although disk surfaces must be very smooth, they occasionally include asperities, or small “hills” and “valleys.” A valley manifests itself as an area in which a disk drive cannot reliably read or write information because the head is too far removed from the surface. A hill can cause more severe problems. An asperity with a height of only 2 µin may cause a head crash that can damage the disk and the head.

Even an asperity of only 1 µin  can cause problems if a head brushes against it. The friction between the asperity and the head causes the head to heat rapidly, typically in 100 to 200 ns. Cooling to ambient temperature can take from 1 to 3 ms. Heating an MR head increases its resistance, which affects its electrical performance.

During the duration of what’s called a thermal asperity, the head’s sense electronics cannot distinguish a logic 1 from a logic 0. Thus, the area of the asperity is lost for the purpose of data storage.

Obviously, asperities can decrease disk-drive manufacturing yields, so failure analysts need to find the actual location on the disk that caused the problem. After they examine the problem area, they can determine how to adjust the manufacturing process to provide better disks. When an operator locates an asperity using a light microscope, he or she must carefully note its position and move the disk to an AFM for further analysis.

Whether caused by an asperity or by mishandling, a read-write head can crash—often called a head slap—into the magnetic media. When it does, it can severely damage the magnetic media. Failure analysts typically use a light microscope to locate the area of impact and then switch to an AFM to look closely at the damaged area (Fig. 1). The AFM will provide topographic information about the crash site that shows the extent and the depth of the damage. In some cases, the impact of the head on the disk may depress the media while in other cases, material also may extend above the impact site. Such an asperity could damage the head.

Detect Magnetic Fields, Too
By switching from an atomic-force to a magnetic-force probe on the SPM, a failure analyst can record the intensity and position of magnetic fields. The field information will show the severity of damage caused to the magnetic film by the head crash (Fig. 2). The pattern of individual bits can disclose stresses in the magnetic film on the disk, damage to the film, and destruction of bit patterns.

These magnetic patterns can help the analyst determine what caused the failure and whether the failure caused irretrievable data loss. Although an MFM scan looks like a topographic map, a magnetic probe operates far enough from the surface so that it does not “sense” topographic effects. Keep in mind that MFM displays show magnetic fields, not heights. The “heights” in an MFM image signify the intensity of a magnetic field.

An MFM has uses outside of failure analysis. Some disk technologies use a keepered surface along with inductive thin-film heads to improve the performance of disk drives. The MFM scans in Figure 3 show the difference between recording on a 70-nm-thick keepered surface and an unkeepered surface. The visualization of the magnetic fields makes analysis of the magnetic characteristics of both types of surfaces fairly easy. The keeper material forms a magnetically conductive film on the disk’s surface to improve the efficiency of the underlying magnetic media.

Manufacturers also use AFMs to examine small bumps specifically formed in landing areas for the disk drive’s read/write head. These small bumps exist in a ring 3–4 mm wide, close to the disk’s central hub. During manufacture of a disk, a laser forms the bumps, that look like small hills or craters.

The bumps provide points at which the head can contact the disk, when the disk is not in use. But because the contact takes place at only a few points, when the disk starts to move, it can easily pull away from the small frictional force exerted by the head on the bumps. To ensure that a head reliably comes away from the bumps, process engineers must carefully control the formation of the bumps, and thus the laser that forms them. And that careful control relies upon height and diameter information supplied by examining the bumps. Bumps typically measure about 0.65 to 0.98 µin  high, and 2 to 10 µin  in diameter.

Although engineers can view bumps through a light microscope, those views yield limited dimensional information. Instead, by using an AFM, they can scan across a field of bumps and get a topographic map that shows the precise dimensions of the bumps. The AFM scan can also provide information about bump spacing, bump roughness, and the surrounding disk surface. Some manufacturing steps may shave away small portions of the bumps, thus reducing their effectiveness. During later stages of production, manufacturers will deposit films of material on the disk surface. A final film of lubricant may be only 20 Å thick. These films may also affect the geometry and dimensions of the bumps formed earlier. A scan by an AFM will help failure analysts determine the source of any processing problems that cause the bumps to be out of spec.

Obviously, not every disk in a production run can undergo inspection by an AFM, so after engineers stabilize a manufacturing process, the production lines rely on other inspection techniques, such as interferometry. The optical techniques do not have enough accuracy for the analytical needs of engineers and failure analysts, however. T&MW

ACKNOWLEDGEMENTS
T&MW appreciates the contributions of Joseph Leigh (Seagate Technologies, Fremont, CA), Raj Rangaraj (Quantum, Shrewsbury, MA), and Albert Wang (ThermoMicroscopes, Sunnyvale, CA), who provided information for this article.

FOOTNOTE
1. Tanghe, Steven, et al., “300 Mbps Read/Write Low-Noise Preamplifier for Magnetoresistive Heads,” MicroNews, Vol. 4, No. 1., First Quarter 1998, IBM Microelectronics, Essex Junction, VT. www.chips.ibm.com/micronews/.

FOR FURTHER READING
Howland, Rebecca, and Lisa Benetar, “A Practical Guide to Scanning Probe Microscopy,” Park Scientific Instruments (now ThermoMicroscopes), Sunnyvale, CA, 1997.

Masi, C.G., “SPMs Tackle Manufacturing Inspection Tasks,” Test & Measurement World, March 1998, pp. 44–48.

Strausser, Yale, and Monte Heaton, “Scanning Probe Microscopy, Technology Overview,” American Laboratory, International Scientific Communications, Shelton, CT. April 1994, pp. 20–28. www.di.com/appnotes/AmLab/AL-SPMMain.html.

Babcock, Ken, “Magnetic Force Microscopy: High Resolution Imaging for Data Storage,” Data Storage, Pennwell Publishing, Nashua, NH, September 1994, pp. 43–49. www.di.com/appnotes/MFM/MFMMain.html