Construction and installation of the U12IR far infrared beamline was completed in 1997, with commissioning taking place in the Fall of 1997. The original spectrometers were a Bruker IFS113v and a custom-built lamellar grating interferometer.

The lamellar grating interferometer. Originally built by Rick Henry and David Tanner at Ohio State University. It served as one of the 1st spectrometers on U12IR and to this day yielded the lowest frequency performance of any IR spectrometer with synchrotron radiation [down to 1 cm-1 or a wavelength of 1 cm]. This photo shows the spectrometer before installatio,along with (from left to right) Ric Lobo (now at ESPCI/CNRS in Paris), David Tanner (U. Florida) and Larry Carr (NSLS).

The Bruker IFS113v interferometer when it was installed at beamline U12IR. This spectrometer was used to test beamsplitter ideas (e.g., silicon as a very wide band beamsplitter for 2 cm-1 and less resolution) and test synchrotron brightness. The need for this particular spectrometer was reduced when the Bruker IFS66v/S became operational at adjacent U10A. This IFS113v was then installed in the BNL Physics department by Chris Homes.

One of the 1st studies performed at U12IR was an investigation of Coherent Synchrotron Radiation (CSR). A few years prior, a beam instability had been noticed at U4IR when the VUV ring was operated with a sizeable charge density in a single electron bunch (for time-resolved IR spectroscopy). The induced noise was a problem for rapid-scan FTIR spectroscopy, but could be avoided by operating at a lower VUV ring beam current. The step-scan lamellar grating interferometer, combined with a beam chopper and lock-in detection, was capable of following the far-infrared intensity even during injection, and a large increase in signal was noticed when the instability was present. NIST had recently reported periodic bursts of microwave radiation from their SURF-II ring, so we quickly realized that the noise was likely due to bursts of coherent synchrotron radiation.

Coherent emission bursts at beamline U12IR, detected directly from the output of a 1.5K Si:B bolometer. The VUV ring was operated with a single electron bunch and at current (time average) of 270 ma, injection energy (745 MeV) and a single RF cavity (52.88 MHz) without bunch stretching.
[R.P.S.M. Lobo, D.B. Tanner and G.L. Carr]

Subsequent studies confirmed that the bursts were due to CSR, and the nature of the instability was investigated throughout 1998 and 1999. The dependence of the instability threshold on synchrotron frequency suggested intrabunch electron-electron interactions were responsible. An effort was made to achieve stable CSR by reducing the bunch charge (beam current) and dispersion in the ring to very low values. But bursting was observed to below 1 ma average beam current even with the synchrotron frequency reduced (from its nominal value of about 11 kHz) to nearly 600 Hz. However, it was noticed that the onset of a burst had slowed to about 200 microseconds, and they appeared more regular intervals (quasi-periodic).

Spectrum analysis of the CSR bursting signal detected using a 1.5K far-IR bolometer. The overall downward trend with increasing frequency is due to the intrinsic thermal time constant of the detector. Notice that structure appears (including "sidebands") as the synchrotron frequency is reduced, indicating more periodic bursting.
[G.L. Carr, S.L. Kramer and J.B. Murphy]

Another spectrum analysis of the CSR bursting signal. The VUV ring dispersion is held fixed (to yield a synchrotron frequency of 1.2 kHz) and the beam current is allowed to drop. All measurements were done with a single bunch.
[G.L. Carr, S.L. Kramer and J.B. Murphy]

Though stable CSR was not achieved at the NSLS, similar efforts at BESSY-II (and more recently ANKA) have demonstrated that "cw" CSR can be produced in more modern storage rings (~500 MHz RF and 3 GeV beam energy) where the electron bunches are intrinsically shorter. This has confirmed aspects of CSR models, which in turn can be used to optimize a storage ring for the purpose of generating CSR as a THz source. The CIRCE ring proposed by Berkeley is such a storage ring source.