The existence of contact binary asteroids was first speculated by
planetary scientist Allan F. Cook in 1971, who sought for potential explanations for the extremely elongated shape of the
Jupiter trojan asteroid
624 Hektor, whose longest axis measures roughly across and is twice as long as its shorter axes according to
light curve measurements.
Astronomers
William K. Hartmann and
Dale P. Cruikshank performed further investigation into Cook's contact binary hypothesis in 1978 and found it to be a plausible explanation for Hektor's elongated shape. They argued that since Hektor is the largest Jupiter trojan, its elongated shape could not have originated from the fragmentation of a larger asteroid. Rather, Hektor is more likely a "compound asteroid" consisting of two similarly-sized primitive asteroids, or
planetesimals, that are in contact with each other as a result of a very low-speed collision. Hartmann theorized in 1979 that Jupiter trojan planetesimals formed close together with similar motions in Jupiter's
Lagrange points, which allowed for low-speed collisions between planetesimals to take place and form contact binaries. The hypothesis of Hektor's contact binary nature contributed to the growing evidence of the existence of
binary asteroids and
asteroid satellites, which were not discovered until the
Galileo spacecraft's flyby of
243 Ida and
Dactyl 1993. Until 1989, contact binary asteroids had only been inferred from the high-amplitude U-shape of their light curves. The first visually confirmed contact binary was the
near-Earth asteroid 4769 Castalia (formerly 1989 PB), whose double-lobed shape was revealed in high-resolution delay-Doppler
radar imaging by the
Arecibo Observatory and
Goldstone Solar System Radar in August 1989. These radar observations were led by
Steven J. Ostro and his team of
radar astronomers, who published the results in 1990. In 1994, Ostro and his colleague
R. Scott Hudson developed and published a three-dimensional shape model of Castalia reconstructed from the 1989 radar images, providing the first radar shape model of a contact binary asteroid. In 1992, the
Kuiper belt was discovered and astronomers subsequently began observing and measuring light curves of Kuiper belt objects (KBOs) to determine their shapes and rotational properties. In 2002–2003, then-graduate student
Scott S. Sheppard and his advisor
David C. Jewitt observed the KBO and
plutino with the
University of Hawaiʻi's
2.24-m telescope at
Mauna Kea, as part of a survey dedicated to measuring the light curves of KBOs. With their results published in 2004, they discovered that exhibits a large, U-shaped light curve amplitude characteristic of contact binaries, providing the first evidence of contact binary KBOs. Sheppard and Jewitt identified additional contact binary candidates from other KBOs known to exhibit large light curve amplitudes, hinting that contact binaries are abundant in the Kuiper belt. The contact binary nature of comets was first suspected after the
Deep Space 1 spacecraft's flyby of
19P/Borrelly in 2001, which revealed a bilobate peanut-shaped nucleus with a thick neck connecting the two lobes. The nucleus of
1P/Halley has also been described as peanut-shaped by researchers in 2004, based on imagery from the
Giotto and
Vega probes in 1986. However, the low bifurcation and thick-necked shapes of both of these comet nuclei made it unclear whether they are truly contact binaries. In 2008, the Arecibo Observatory imaged the
Halley-type comet 8P/Tuttle in radar and revealed a highly bifurcated nucleus consisting of two distinct spheroidal lobes, providing the first unambiguous evidence of a contact binary
comet nucleus. Later radar imaging and spacecraft exploration of the
Jupiter family comet 103P/Hartley in 2010 also revealed a thick-necked, peanut-shaped nucleus similar to 19P/Borelly. By that time, half of the comets that have been imaged in detail were known to be bilobate, which implied that contact binaries in the comet population are similarly abundant as contact binaries in other minor planet populations. == Formation and evolution ==