Everyone knows what bar codes are. The zebra-striped images, gummed labels or tags are on just about every item in every store. They prevent pricing errors and speed things up at the checkout counter, so there's no question that they, or rather, the inventory control system they represent, benefit the consumer. Bar codes are the visible part of the globally accepted Uniform Product Code (UPC) system used to track the movement of goods from manufacture to retail sale. They also track the movement of things like massive international freight containers, checked airline luggage and blood samples, to name just three examples.

At least they do so for now, because bar codes are obsolescent. Useful as they are, the little striped labels have a number of limitations, the most serious of which is that they require line-of-sight visibility in order to be read by an optical scanner. Even then, a little dirt can impair their functionality. They often require a human interface, like a supermarket clerk, and they're limited with respect to the amount of data they can contain. That information is usually the UPC number that identifies the item and its manufacturer. If more information is needed, one or more additional tags are applied. Finally, the tags are entirely passive, meaning they can't provide anything other than static facts about whatever they're affixed to: identity, date of manufacture, shipping date, destination, etc.

The technology began to change about 25 years ago with the emergence of a new type of tag that can be read using radio-frequency (RF) waves. Because RF penetrates most materials, the new tags don't require line-of-sight visibility. They are currently used by a few thousand companies worldwide, but they will eventually outnumber their optical counterparts. More than a billion of these RFID (for RF identification) tags or transponders were deployed globally as of mid-2005,¹ but that's just a tiny fraction of what's to come. Copper, which seems never to be far from new technology, is one of their key ingredients.

How It Works

Here's how it works: a typical passive RFID tag (the most common type, similar to a bar code label in function) consists of a microchip about the size of a grain of sand connected to a tiny antenna, plus some form of packaging to protect the device and attach it to the product being tagged. The reader used to interrogate the RFID emits a signal that inductively couples with the tag's antenna, momentarily creating enough electrical current to energize the microchip. The chip responds by modulating the reflected RF signal in such a way as to upload its digitally coded information back to the reader, from where it is sent to a computer for processing and storage. The process takes a few milliseconds and can be conducted at distances ranging from a few inches to about 20 feet.

Tags can be small enough to be embedded into such things as smart credit cards or encapsulated in glass beads for implantation in pets. Larger versions are used in materials handling systems. The microchip might contain only identification in the form of an Electronic Product Code (EPC, analogous to the UPC but more extensive), but, because memory can be expanded to more than 125 Kbytes, it can also contain such information as manufacturing location and date, shipping instructions, customer identification, invoicing data and even an individual item's serial number.

The maximum useful interrogation distance, or read range, depends on a number of factors. Frequency is very important, but so is the antenna and what it's made from. Copper, which produces the longest read ranges, has been used for RFID antennas from the very start. Early (1980s-vintage) tags operated at relatively low frequencies, 125 kHz to 148 kHz, not far below the long-wave AM radio band. Antennas were made from copper coils, either wound from wire or etched out of foil in much the same way conductors are formed on printed circuit boards (PCBs). These are expensive processes, but they provide RFIDs with a read-range up to several feet.

RFID tags began finding applications in access control, product identification, automobile immobilization and livestock identification; however, they were far too expensive to displace bar codes for uses such as inventory control, even when bought in huge quantities.

Cost is less of a barrier in the case of active RFID tags, which can run $50 to $70 apiece or more. Unlike their passive cousins, these devices contain a battery to power the microchip. They also contain sensors to provide stored or real-time information about the tagged object. For example, an active tag might report the temperature history of a carton of hamburgers, or whether there had been an attempt to breach security in a shipping container. NASA is developing active structures capable of logging exposure during space flight. Taking the technology one step farther, the agency's scientists have developed a hairdryer-sized gun capable of applying thin metal coatings, including copper, on just about anything. Suitably masked, such coatings could easily be made in the form of an RFID antenna.²

The 1990s saw a push toward lower tag costs through development of systems operating at high frequencies (HF), the most common of which being 13.56 MHz. This change offered higher data transfer rates for speedier operation. Tags could be fitted with antennas that were printed onto substrates as a 25- to 35-µm-thick layer of silver-containing conductive ink, a cheaper process than etching copper.³ Costing as little as 20 cents, the tags found applications as mundane as tracking library books and identifying laundry. It appeared for a while that copper's role in RFIDs would diminish in favor of ink, at least for tags operating at this frequency. On the other hand, the 13.56 MHz frequency itself has some technical limitations, and some industry observers speculated that these shortcomings, coupled with the development of ultra-high frequency (UHF) 915-MHz and microwave frequency 2.45-GHz tags plus improvements in the manufacture of low-frequency 125-kHz tags might relegate the 13.56 MHz spectrum to niche applications. That did not happen; in fact, some major users, including the international shipping firm DHL, are currently committed to the frequency. Meanwhile, 13.56-MHz tags with soldered copper antennas continued to be favored for their unsurpassed reliability.

Retailer Spurs Cost Reductions

UHF tags operating between 868 MHz and 956 MHz gained popularity in the early 2000s; 915 MHz has emerged as the most favored frequency in the spectrum. In addition to their reasonable cost, UHF systems offer the advantage, called non-collision, of reading large quantities of tags simultaneously at distances up to 20 ft. Visualize pallet-loads of cartons full of thousands of widgets being read accurately as they pass an interrogation portal at 10 ft per second and you can appreciate the cost-savings such a system enables.

RFID technology, and UHF tags in particular, got a major shot in the arm in June 2003 with the announcement by Wal-Mart that the world's leading retailer would require its 100 largest suppliers to incorporate RFID technology beginning in January 2005. Wal-Mart used its clout to overcome the RFID industry's longstanding dilemma: creating a large enough user base to reduce tag prices to the point at which they can be applied to low-cost products. Wal-Mart's move makes good business sense: by accelerating adoption of RFID, the company might save itself $1.5 to $1.8 billion per year in inventory costs.⁴ Understandably, several more retailers are now either seriously looking into RFID or have committed to it. The U.S. Department of Defense is helping, too, by issuing guidelines to encourage its suppliers to adopt RFID technology beginning at the end of 2005.

But the cost of individual tags is still the major hurdle standing between passive RFID technology and the volumes enjoyed by bar codes. So-called chipless tags might be one emerging solution. These tags contain no microchip, but their antennas, which can be copper, can reflect an RF signal in which (currently) up to 24 data bits can be encoded. That's far less than the 96-bit minimum contained in a conventional RFID tag, but it might be sufficient to edge out some bar-code applications. Costs of chipless tags are estimated to drop to well less than a cent in large quantities.

Lower-Cost Copper Antenna Technology

Various industry sources talk of five-cent RFID tags as the as-yet-unattained Holy Grail. Those numbers likely won't be gained through sheer volume alone; they will require advances in technology, and copper is, again, a key part of that technology.

The United Kingdom-based PCB technology development firm TDAO recently announced a process by which copper antennas can be electrodeposited (plated) precisely onto substrates at extremely high speeds.⁵ The technology is reportedly cost-competitive with conductive-ink printing while offering denser, higher-conductivity and, therefore, higher quality copper antennas. The company will have the capacity to produce several billion 13.56-MHz HF antennas per year on its prototype 1-meter-wide line by the end of 2005. The company plans to develop higher-capacity equipment in 2006.

The ingenious TDAO process makes use of a very thin (and therefore inexpensive) layer of conductive ink. Too thin for use as an antenna, the ink acts as a target for the electroplated copper, which is deposited from an inexpensive copper sulfate electrolyte. The result: a 100% dense metallic copper antenna that provides all the performance of the etched-foil product but at a fraction of the cost.

A Dutch plating equipment supplier, MECO, offers a similar process, which it claims can reduce the cost of HF antennas by 60% to 75%, and by even more for UHF antennas, which are typically thinner. As with the TDAO process, the final antenna is in the form of fully dense copper.

Qinetiq, another British innovator, offers a somewhat different scheme. In this case, copper is deposited on a thin ink layer by a process known as electroless plating, which is a chemical rather than an electrochemical deposition process. In it, copper plates out only onto areas preprinted with conductive ink. The electroless-plated copper layer is very thin, and a subsequent electroplating step is necessary to bring the antenna to full thickness. The process sounds complex, but MECO claims it produces antennas at about one-half the cost of conventional manufacture. However, the process' most important feature might be its ability to grow copper between the antenna and printed contacts on the microchip, thereby eliminating a separate assembly operation.