Following Penzias and Wilson's great discovery, it was clear that measuring the microwave background near its peak intensity at a wavelength of 2 mm was the crucial observational test to demonstrate the blackbody spectrum of the radiation and its cosmological origin. Most of the energy of a 3-K black-body lies in the vicinity of 2 mm, and because the energy in the CMBR is enormous on a cosmic scale - 100 times that of starlight when averaged over the great voids between galaxies - that is the observation which constituted the dagger at the throat of the steady state universe. Because the opacity of the terrestrial atmosphere increases greatly at short wavelengths, it seemed likely that a spacecraft observation was required, with all the expense, difficulty, and delay which that was likely to entail.
I had recently designed a small radio telescope for the Nimbus satellite to study the thermal emission of the Earth at a wavelength of 2 cm (chosen because that is a band where the contrast between water and land is large, and icebergs are readily distinguished against the surrounding ocean). I was therefore well aware of the technical and political problems - and the frustrations - which a spacecraft measurement imposed. I discussed these from time to time with my colleague William Hoffman at the Institute for Space Studies in New York City, who was working in collaboration with Neville Woolf on a small balloon-borne cryogenic telescope to conduct the first survey of the Milky Way in the far infrared. That instrument, which went on to discover how remarkably rich the Galaxy is in the far infrared, was the forerunner of Infrared Astronomical Satellite (IRAS) and the other far-infrared telescopes which have had such a large impact on space astronomy.
Woolf had been in New York working with Hoffman but had recently gone to the University of Texas, where he remained a member of our informal discussion group, and he was aware of my interest in the microwave background. From there in September 1965 he wrote me a letter which set me back, because it suggested that I might be wasting my time. "On a theoretical approach," he wrote, "I wonder what can be made from the case of the missing CN lines. Lines arising from a level only 3 cm-1 above the ground state are absent." This came as a severe shock and a disappointment, because it implied that there might be no background radiation at all in the millimeter-wave band, and that the Penzias and Wilson radiation - whatever it was - was not the faint blackbody remnant of the primeval fireball predicted by Gamow. An attempt to observe directly the short wavelength background could be a significant waste of time and money.
So I decided to look into the question of interstellar CN and the mechanisms which excite it in the interstellar gas. It didn't take long to discover that this widely studied radical, readily observed on the Sun, in comets, and in laboratory discharges and flames, was not stationary in the interstellar gas at all, but was rotating instead by just the amount expected from excitation by 3-K blackbody radiation. Observed as an extremely faint satellite, the R(1) line, to the R(0) line from the rotational ground state, this seemingly insignificant fact made almost no impression on astronomers when it was first observed by McKellar in 1940 as a barely perceptible absorption line on a high-resolution Mt. Wilson spectrum of the second-magnitude star Zeta Ophiuchi. But Gerhard Herzberg, the eminent molecular spec-troscopist and a sharp-eyed observer who missed little, was aware of it - to the point of citing it on the penultimate page of his well-known monograph Spectra of Diatomic Molecules, whose first edition appeared in May 1950. There Herzberg quotes a rotational temperature of 2.3 K, but proceeds to say that this excitation has "... only a very restricted meaning."
This conclusion was far from unreasonable at the time - it is probably what most astronomers and spectroscopists would have said had they taken the trouble to consider the matter at all. In 1940 only three molecules were known in the interstellar gas: CH, CH+, and CN. As the heaviest of the three, and the one with the lowest frequency rotational transitions, CN was the most susceptible to excitation by purely local processes such as collisions with H atoms or resonant scattering (fluorescence) of starlight. Herzberg therefore reasoned that CN was the first interstellar molecule which one might expect to find rotationally excited in space, and its excitation was hardly surprising and had no general significance. It presumably varied from interstellar cloud to cloud, in response to the local density, intensity of starlight, etc.
But with the scant interstellar CN data which existed at that time and the poor signal-to-noise which characterized all of it, variation from cloud to cloud was very difficult to demonstrate. Astronomers would have been puzzled to discover that CN everywhere was excited by the same amount: just under 3 K. Was Herzberg right? was the crucial question.
About this time I took on for his doctoral research a versatile and energetic graduate student in the Columbia Physics Department, John F. Clauser, and together we undertook to do two things to see if CN really served as a good thermometer for the CMBR. The first was theoretical: to see if local processes were fast enough to explain the CN excitation. The second was observational: to obtain better CN data to better measure the CN excitation and the CMBR temperature, and to see whether the excitation was constant from cloud to cloud, as predicted by a universal mechanism.
Our theoretical calculations were hampered at the outset by the fact that the CN electric dipole moment - the parameter that determines the rate of rotational excitation by background photons and how tightly CN is coupled to the CMBR - had not been measured. So it had to be obtained indirectly. In our first paper, Clauser and I argued that the small value for the dipole moment of the first excited electronic state of CN, which had been obtained from pressure-broadening in a flame, implied that the dipole moment of the ground state was probably substantial, at least 1 Debye (in the standard unit of molecular dipole moments, the Debye, 10-18 esu), making CN a sensitive thermometer for the CMBR. That assumption turned out to be correct. Thanks to a communication from George Field, who we discovered was also pondering the question of the excitation of interstellar CN, we then learned that the Belgian astronomer Claude Arpigny had recently deduced the dipole moment of CN to be 1.1 D from cometary spectra, and that is what we adopted for our calculations. A few years later Thompson and Dalby observed the optical Stark effect of CN in the laboratory - a difficult experiment - and with it measured directly the CN dipole, and found it to be even larger: 1.45 D. So our indirect estimate turned out to be conservative: CN was even more tightly coupled to the CMBR than we had assumed, and was an even better radiative thermometer than we had supposed.
Clauser and I soon calculated that if the CN was located in a normal HI region, as generally assumed, the excitation by local processes, in particular collisions with hydrogen atoms and fluorescent excitation by background starlight, was quite slow compared to that from the CMBR, and it was therefore the CMBR which largely determined the excitation. We found that in an
HII region the situation was less clear-cut: collisions with charged particles, slow protons in particular, could be a significant source of CN excitation; but there was then - as today - little evidence that interstellar CN is found in HII regions, so that possibility could be discounted. Our conclusion was therefore, contrary to what Herzberg had assumed, that the excitation of CN was the work of the CMBR, and the rotational temperature of CN was a direct measure of the temperature of the background at the wavelength of the 1-0 rotational transition, 2.64 mm - quite close to the peak of the 3-K blackbody curve, which is the reason why our measurement was important. It implied that the enormous amount of radiant energy locked up in the CMBR which had escaped detection by the low-frequency measurements of Penzias and Wilson and other radio astronomers was really there, and could not be swept under the rug as the proponents of the steady state theory might have liked.
Our attempt to show that the CN excitation was constant from star to star was somewhat disappointing, because in the 1960s the number of bright stars with interstellar CN absorption was very limited, consisting of only two really good examples: ( Oph, the second-magnitude star studied by McKellar, and ( Per, a similar star in the opposite part of the Milky Way. We were nonetheless able to conclude from a total of ten stars with some evidence of interstellar CN that all the available data were consistent with invariant CN excitation, and the existence of background radiation at a temperature of 3 K or somewhat less. Today with modern echelle spectrographs and CCD detectors, one could demonstrate the constancy of CN excitation much better than in the 1960s, but with the remarkably accurate direct measurement of the CMBR spectrum by COBE, the question now is largely of historical interest.
Over the next few years we made an effort to obtain a really good measurement of the CMBR with CN and the other molecules then known to exist in the interstellar gas. This was well before the flood of interstellar molecule discoveries, which began with the discovery of ammonia and water by Townes and Welch and coworkers at Berkeley in 1968-1969. Radio lines of OH had been identified in 1963, but OH optical lines analogous to those of CN do not exist, and this widely distributed molecule unfortunately does not serve as a useful radiative thermometer for the background. Of the more than 130 molecules now known to exist in the interstellar gas or circumstel-lar shells, it is remarkable that none surpasses CN as a thermometer for the CMBR - or even comes close.
For a radio astronomer, obtaining the necessary time in the 1960s on a large optical telescope with a fast high-resolution spectrograph was far from easy. To get started, Clauser and I succeeded in begging time on the large McMath solar telescope at Kitt Peak, which had a very high-resolution spectrograph and had been used for a small amount of interstellar work, and with it we succeeded in obtaining a few rather ragged spectra of Z Oph showing interstellar CN. These yielded a rotational temperature of about 3 K, but our data really represented little or no improvement over McKellar's pre-war observations at Mt. Wilson. It became clear after discussions with optical observers who worked at high spectral resolution that the instrument to use for a significant improvement was the Lick 120-inch telescope, which thanks to George Herbig had a very high-resolution Coude spectrograph that was fast and efficient.
To obtain a substantial block of time on this telescope as a visiting faculty member of the University of California, I moved with my wife and two children to Berkeley for the spring quarter of 1968, and was granted then and over the summer a substantial block of time on the 120-inch telescope for the CN project by Albert Whitford, the Lick Director. With another New York student, Victor Bortolot, I ultimately obtained over 30 very high-quality Coude spectra on IIao emulsions, carefully baked to enhance sensitivity, at the highest available resolution of the spectrograph, 1.2 A/mm; most were 5 mm wide to store a large amount of information on each exposure, which typically required several hours. Back in New York we digitized these spectra and added them together numerically to obtain the interstellar spectrum of Z Oph with a signal-to-noise not previously achieved. I was fortunate to be tutored in the art of high-resolution Coude spectroscopy by two masters, George Herbig, then on the faculty at Santa Cruz, and Gene Harlan, a fastidious and exacting member of the Lick technical staff.
During one of our Lick runs we were paid an impromptu visit by George Mueller, director of the Apollo Program, and Wernher von Braun, director of the Marshall Space Flight Center and the chief architect of the Saturn V launch vehicle developed for the Moon landing. They were contemplating a telescope of the Lick class in orbit and wanted to see what one actually looked like. As a prototype for a space telescope, the Lick 120-inch with its long focal ratio was a bizarre choice, and one is thankful that this visit did not sour NASA on a big telescope in space for good. Mueller soon left, but von Braun, who knew something about optics and had an affection for telescopes, decided to stay the night. I put him to work in the bowels of the Lick Coude guiding the telescope, keeping the light from the star streaming down the entrance aperture of the spectrograph. He was an interesting companion for a long observing night, and a good talker. It was not long before this - about 1960 - that von Braun published his memoir I Aim at the Stars, which a well-known comedian (Mort Sahl) said might have been better called I Aim at the Stars, and Sometimes Hit London, and it was about the time Norman Mailer in his book Of a Fire on the Moon described von Braun as looking like "the head waiter in the largest hofbrau house in Heaven." It was an unusual evening at the Coude focus of the 120-inch.
Our synthesized spectrum represented well over 100 h of 120-inch observation. With it we determined the rotational temperature of CN toward Z Oph to be 2.99 ± 0.06 K, which with small corrections for optical depth and collisional excitation yielded 2.78 ± 0.10K for the temperature of the CMBR at A = 2.64 mm. It is gratifying that this result obtained by photographic spectrophotometry in 1972 is within 0.6a of the COBE temperature published 18 years later.
Our synthesized final spectrum covered not only the violet band of CN, but also, as Figure 4.3 shows, the stronger interstellar bands of CH and CH+, and it yielded also a marginal detection of 13CH+, which furnished the first observational evidence for carbon-13 in the interstellar gas. The 13C/12C ratio, which we obtained, was consistent with the terrestrial value, 1/89, and with the many measurements later made of this important isotopic ratio.
Several years later I took on as a postdoc in radio astronomy John Mather, a brilliant student of Paul Richards. For a research topic John took up the recently discovered SiO masers, but, true to his training in Richards' laboratory, his real interest was in the microwave background and the challenge of its short wavelength spectrum. The first proposal for a Cosmological Background Radiation Satellite was written in my office in New York. Already present at our first meeting to draft a proposal to NASA for a spacecraft experiment were some of the key COBE players, including Weiss, Wilkinson, Hauser, and Silverberg. Mather was so good at both the political and the technical requirements of this enterprise that when he later took a job at the Goddard Space Flight Center, the project followed him with my support. The discovery of polyatomic molecules in space was underway, and that promised the kind of science which was closest to my heart.
My excursion into optical astronomy was short, but it left an indelible impression. After a many-hour exposure, the thrill of holding up a developed photographic plate in the darkroom to the light and seeing the faint, barely perceptible absorption line of excited CN, knowing that it was a fingerprint of the universal radiation filling all space, once as brilliant as the surface of the Sun, was an aesthetic and intellectual pleasure which I have
Was this article helpful?