Broadband FTMW

Click on above pattern for explanation

We have recently developed a broadband Fourier transform microwave (FTMW) spectrometer. By dramatically increasing the bandwidth of spectrum we probe at one time, we have been able to fundamentally change the way a microwave spectrum is acquired. Instead of the traditional way of acquiring a spectrum one narrow (~1 MHz) step at a time, we can now shine a broadband MW pulse on large (up to 11 GHz) regions of the spectrum and acquire the entire spectrum of interest in one or two steps.

The technical breakthrough was realized by use of an arbitrary waveform generator and its ability to perform linear frequency sweeps in short times (1 GHz in 2 nsec). The fast sweep rates of the waveform generator have an advantage over the microwave synthesizers traditionally used in rotational spectroscopy that are limited to millisecond sweep times. Linear frequency sweeps provide fast adiabatic passage as the excitation method. In comparison to transform limited pulses where the signal is proportional to the inverse of the bandwidth, the linear sweep technique provides rotational signals that scale as the inverse of the square root of the excitation bandwidth. This is shown graphically in Figure 1.

In order to acquire broadband spectra with a satisfactory signal to noise ratio, it is imperative to be able to average the molecular free induction decay (FID) in the time domain. The arbitrary waveform generator allows us this capability, as well as complete phase control of the molecular signal.

The experimental spectrum of suprane (black spectrum), a medical anesthetic, is compared to the calculated spectrum (red, blue, and green spectra, representing A, B, and C type transitions respectively) in Figure 2. The 11 GHz spectrum was recorded in one step, averaging 10,000 shots (45 minutes of acquisition time). The relative intensities of the experimental spectrum agree very well with the calculated intensities at 1 K. Click on the spectrum to see a zoom in of the area Q-branch in the pink boxed area.

The pure rotational spectrum of epifluorohydrin is shown in Figure 3 along with the calculated intensities (dipole moments found by ab initio calculation). The 10 GHz spectrum was measured in 2 steps, of 10000 averages each (less than one hour of data acquisition time). For an impressive display of the sensitivity of our new spectrometer for species of low natural abundance (13C species) as well as species with weak dipole moments (other conformations of epifluorohydrin), click on Figure 3.

We are also able to perform MW-MW double resonance techniques in order to probe connectivity of MW resonances in pure rotational spectra. In Figure 4, a section of the pure rotational spectrum of epifluorohydrin is compared to the same section of spectrum when the 414-404 rotational transition (10439.57 MHz) is pumped after the initial MW polarizing pulse, but before data acquisition. In this case, the driving of the 414-404 MW transition completely destroys the coherence of a the 413-404 transition and the 404-313 transition. Furthermore, this MW pulse decreases the intensity of the 321-414 MW transition. In order to see the pulse sequence used in this experiment, click on Figure 4.

We have also been able to improve some of our dynamic rotational spectroscopy measurements we previously measured with our high resolution spectrometer. Figure 5 shows the dynamic rotational spectrum of vibrationally excited (two quanta in the acetylenic CH stretch) trifluoropropyne. The same measurement is shown for both the high resolution and the broadband spectrometer. It is evident that not only does the broadband spectrometer reproduce the overall lineshape measured by the high resolution spectrometer, but it also improves the signal to noise ratio. Furthermore, the broadband spectrometer is better suited for measuring the high density of resonances seen in this spectrum, which is important as we move toward measuring the dynamics of molecules with denser spectra. Finally, the previous high resolution spectrum required over one hour of data acquisition, while the new broadband FTMW spectrometer acquired the spectrum in thirty minutes. This difference in acquisition time becomes much more dramatic when measuring larger regions of frequency spectrum. For a better explanation of dynamic rotational spectroscopy, see the area of the webpage dedicated to dynamic rotational spectroscopy.

The double resonance measurement shown in Figure 5 is actually a probe of multiple quantum states, since our 0.02 cm-1 bandwidth IR laser is sufficiently broad to excite multiple quantum states with each laser pulse. In order to perform a quantum-state resolved measurement, we must do an IR-MW-MW triple resonance measurement. An example of a triple resonance measurement performed with our new broadband FTMW spectrometer is shown for 4-fluorobutyne in Figure 6. In this case, the R(2) rovibrational transition of 4-fluorobutyne was pumped with our IR laser to populate the v=1, J=3 rovibrational level. Then a narrowband MW pulse (our broadband FTMW spectrometer is also capable of producing narrow band MW pulses) was used to select a single quantum state in the v=1, J=2 rovibrational level. Finally, we used a broadband (in this case 6.5 GHz) MW pulse (linear frequency sweep) to probe molecular transitions from the v=1, J=2 level into both the v=1, J=3 and the v=1, J=1 rovibrational levels. The spectrum shown below was acquired in two steps, the first one covering the 7.5 - 14 GHz range, and the second one covering the 11.5 to 18 GHz range of the spectrum. The spectrum includes only those resonance lines which originated out of the v=1, J=2 level (pure rotational transitions as well as double resonance transitions have been removed). Therefore, this is a quantum state resolved spectrum. We were able to use an overall lineshape analysis of this spectrum to find the isomerization lifetime of 4-fluorobutyne at 3330 cm-1 to 37 ps (RRKM theory predicts a lifetime of 1.4 ps). This study is similar to the isomerization study performed on pent-1-en-4-yne using our narrow band FTMW spectrometer described in the dynamic rotational spectroscopy section of this website.

Due to our fast acquisition time, we can now measure changes in the microwave spectrum as the IR laser is actively scanned. Fig. 7 shows the MW spectrum of vibrationally excited cyclopropanecarboxaldehyde as the IR laser is scanned from 2818 - 2825 cm-1. All pure rotational transitions have been removed from the data for simplification. In this region of the spectrum, only the cis conformer is being pumped with the IR laser, so evidence of MW transitions in the region characteristic of the trans conformer gives evidence of IR laser induced isomerization. The overall lineshape can be used to measure the isomerization rate, as in 4-fluorobutyne above.