1. Millimeter-wave spectroscopy of formyl azide (HC(O)N3)
Nicholas A. Walters, Brent K. Amberger,
Brian J. Esselman, R. Claude Woods, Robert J. McMahon
University of Wisconsin
June 26, 2015

2. Formyl azide (HC(O)N3) No rotational data in the chemical literature
Spectroscopic findings include IR, UV and NMR spectra, as well as crystallographic data1,2
Structurally similar to HN3 and C(O)(N3)2
Banert and Hagedorn are common to each paper
Solution phase studies via IR, UV and NMR spectroscopies in CCl4 – ref 1 from tbo
Interesting to us because it shares structural similarities to molecules like HN3 and carbonyl diazide, both of which have been studied by our group.
According to calculations at CCSD(T) level of theory with ANO1 basis set, the two isomeric forms of formyl azide (syn or anti configuration of the C=O and N3 groups with respect to the C-Na bond) which we expect to see in the gas phase differ in energy by 2.8 kcal/mol
Cs symmetry
Syn-isomer
Anti-isomer
Relative E (kcal/mol)
(CCSD(T)/ANO1)
2.8
1. Banert, K. et al. Angew. Chem. Int. Ed. 2012, 51,
2. Zeng, X. et al. Angew. Chem. Int. Ed. 2013, 52,

3. Synthesis Highly unstable molecule at room temperature
t1/2~ 20 min in CDCl t1/2~ 2 hours in gas-phase
Sample frozen in liquid nitrogen
Outlines the simplest version of a Curtius rearrangement
LN2 trap to crudely remove water, ethylene glycol and dry ice to remove more water and allow formyl azide through
Dry ice-ethylene glycol trap used to remove water from sample – allowed FA through, while other traps didn’t?
Banert, K. et al. Angew. Chem. Int. Ed. 2012, 51,

4. Spectrum collection ~10 mTorr 250-360 GHz Room temperature
Mention about 8 hours per segment of about GHz
Sample through series of traps before reaching sample chamber to
Sample pressure not super precise as we bleed in gas evolved from the sample initially containing water vapor as well
Mention how some spectral range segments aren’t as strong as others due to sample either being “tired” toward the end of data collection or the sample decaying more rapidly in some segments
We put up to 20 mTorr of sample into our sample chamber of 3 m in length. We then sweep the frequency of the source ( GHz) while detecting the intensity of transmitted radiation, resulting in our absorption spectrum which we can use to analyze our molecule.

5. Syn-isomer predicted spectra
μa (D)
μb (D)
μc (D)
0.224
1.547
Syn-isomer ground state
Ground states
Note how b-type transitions dominate in syn, a-type transitions dominate in anti
CCSD(T)/ANO1

6. Syn-isomer spectral characteristics
b-type Q-branch transitions
Experiment
k-1 1413 transitions
Ground state fit
Kprolate 14<-13 series of ground state
First line in series is J 14, note the turnaround around J 16
Showing the anatomy of the spectral features while showing the good agreement of the fit with experiment
J=19
J=18
J=17
J=16
J=14
J=15

7. Syn-isomer spectral characteristics - continued
b-type R-branch transitions
Experiment
k-1 10 transitions
Ground state fit
Kprolate 10 transitions
Point out kprolate 10, 21, 32 and 43 transitions and note the general progression seen within a clump of R-branch transitions.
Can take a closer look at these transitions to see a repeating “clump” of transitions
k-1 21
k-1 32
k-1 34
k-1 43
k-1 45
k-1 54
k-1 10

8. Fit of syn-isomer ground state
CCSD(T)/ANO1
Experimental constants
A (MHz)
B (MHz)
C (MHz)
ΔJ (kHz)
ΔJK (kHz)
ΔK (kHz)
δJ (kHz)
δK (kHz)
A (MHz)
(92)
B (MHz)
(28)
C (MHz)
(28)
ΔJ (kHz)
(16)
ΔJK (kHz)
(48)
ΔK (kHz)
(65)
δJ (kHz)
(28)
δK (kHz)
(98)
ΦJ (Hz)
(32)
ΦJK (Hz)
(21)
ΦKJ (Hz)
(17)
ΦK (Hz)
4.448(13)
φJ (Hz)
(74)
φJK (Hz)
(28)
φK (Hz)
1.9886(38)
LJJK (mHz)
(12)
LJK (mHz)
(25)
LKKJ (mHz)
0.0739(12)
lJ (mHz)
(50) n
1236
Again, mention types of transitions included in fit
May just take the time to mention low-intensity a-type R-branch transitions which are included in fit as well (only a handful of these included in fits – not used in higher energy vibrationally excited states).
Constants came from fak3ab.asf – could include more distortion terms to decrease fit deviation, however adding in more marginally changes the deviation

9. Syn-isomer excited vibrational states
cm-1
2ν12 + ν9
cm-1
2ν9 + ν12
cm-1
ν11
cm-1
3ν9
cm-1
2ν12
cm-1 ν8
cm-1
ν9 + ν12
Now began to assign transitions belonging to higher energy vibrationally excited states
cm-1
2ν9
cm-1
ν12
cm-1 ν9
0 cm-1
ground

10. Syn-isomer: ν9 ( cm-1)
Loomis Wood plot based on calculations for ν9 (centered on bQ k-1 1413 series)
A (MHz)
(13)
B (MHz)
(43)
C (MHz)
(51)
ΔJ (kHz)
(26)
ΔJK (kHz)
(57)
ΔK (kHz)
(10)
δJ (kHz)
(31)
δK (kHz)
(21)
ΦJ (Hz)
(47)
ΦJK (Hz)
(59)
ΦKJ (Hz)
(22)
ΦK (Hz)
4.968(23)
φJ (Hz)
(85)
φJK (Hz)
(65)
φK (Hz)
2.2411(94) n
955
Level of theory for calculation of alpha values should be put in somewhere – CCSD(T)/ANO1, same as previous calculations
LW plot centered on b-type Q-branch series of kprolate 1413 – shows how useful initial predictions from computations were in finding vibrational states of the syn isomer, as this LW plot was generated just from computed rotation-vibration interaction terms
A’ (totally symmetric)
Calculation
Experiment αA αB
-6.540
-6.455 αC
0.605
0.617

11. Syn-isomer: ν12 ( cm-1)
Loomis Wood plot based on calculations for ν12 (centered on bQ k-1 1211 series)
A (MHz)
(17)
B (MHz)
(65)
C (MHz)
(77)
ΔJ (kHz)
(40)
ΔJK (kHz)
(83)
ΔK (kHz)
(14)
δJ (kHz)
(49)
δK (kHz)
9.3096(42)
ΦJ (Hz)
(74)
ΦJK (Hz)
0.0041(12)
ΦKJ (Hz)
(39)
ΦK (Hz)
4.662(30)
φJ (Hz)
(14)
φJK (Hz)
0.0286(12)
φK (Hz)
2.2411(94) n
785
Another example of the same Q-branch series predicted well again for v12.
A’’ (anti-symmetric)
Calculation
Experiment αA αB
9.867
10.203 αC
0.717
0.786

12. Syn-isomer: 2ν9 ( cm-1)
Loomis Wood plot based on calculations for ν12 (centered on bR k-1 21 series)
A (MHz)
(17)
B (MHz)
(60)
C (MHz)
(71)
ΔJ (kHz)
(35)
ΔJK (kHz)
(90)
ΔK (kHz)
(14)
δJ (kHz)
(57)
δK (kHz)
(51)
ΦJ (Hz)
(62)
ΦJK (Hz)
0.0140(15)
ΦKJ (Hz)
(51)
ΦK (Hz)
5.207(31)
φJ (Hz)
(14)
φJK (Hz)
0.0233(15)
φK (Hz)
2.657(29) n
757
Mention how series to the left are lower-energy vibrational states’ R-branch transitions – likely ground and v9 or v12
A’ (totally symmetric)
Calculation
Experiment αA αB αC
1.210
1.064

13. Syn-isomer: ν9 + ν12 (420.30 cm-1) Syn-isomer: ν11 (582.60 cm-1)
A (MHz)
(16)
B (MHz)
(61)
C (MHz)
(70)
ΔJ (kHz)
(36)
ΔJK (kHz)
(91)
ΔK (kHz)
(14)
δJ (kHz)
(62)
δK (kHz)
(51)
ΦJ (Hz)
(66)
ΦJK (Hz)
0.0088(15)
ΦKJ (Hz)
(52)
ΦK (Hz)
5.021(30)
φJ (Hz)
(16)
φJK (Hz)
0.0233(15)
φK (Hz)
2.657(29) n
672
A (MHz)
(68)
B (MHz)
(26)
C (MHz)
(30)
ΔJ (kHz)
4.4781(15)
ΔJK (kHz)
(51)
ΔK (kHz)
(57)
δJ (kHz)
(32)
δK (kHz)
9.696(31)
ΦJ (Hz)
(30)
ΦJK (Hz)
0.0281(81)
ΦKJ (Hz)
-1.197(30)
ΦK (Hz)
4.97(12)
φJ (Hz)
(84)
φJK (Hz)
φK (Hz)
2.77(15) n
493
Fit deviation increases marginally with more distortion terms, however the distortion term values themselves have larger errors associated with them
Using favib7and8b.v11 is A’’ (anti-symmetric)
Calculation
Experiment αA αB
3.327
4.386 αC
1.322
1.549
Calculation
Experiment αA αB
2.704
2.943 αC
0.772
0.732

14. Excited vibrational states of the syn-isomer
cm-1
2ν12 + ν9
cm-1
2ν9 + ν12
cm-1
ν11
cm-1
3ν9
Potential for Fermi resonance
cm-1
2ν12
cm-1 ν8
cm-1
ν9 + ν12
Looking at all of these vibrational states in the hopes of greatly simplifying search for anti-isomer by removing transitions from the experimental spectrum
Summary of current fits obtained for vibrationally excited states
Talk about 3v9, 2v12, v8 and potential perturbations between these states
V9 is A’ (totally symmetric), so 3v9 is totally symmetric (A’)
V8 is totally symmetric (A’)
V12 is A’’ (anti-symmetric), so 2v12 is A’ (totally symmetric)
Think Cs point group A’ 1 1
A’’ 1 -1
cm-1
2ν9
cm-1
ν12
cm-1 ν9
0 cm-1
ground

15. Anti-isomer predicted spectra
Syn-isomer
μa (D)
μb (D)
μc (D)
2.067
1.502
μa (D)
μb (D)
μc (D)
0.224
1.547
Anti-isomer
Syn-isomer
Relative E (kcal/mol)
2.8
Anti-isomer ground state
Ground state
Mention how transition spacing is different here than in the syn-isomer spectra, so finding the transitions using LW plots should be a bit easier
Note how b-type transitions dominate in syn, a-type transitions dominate in anti
Mention size of dipole moment being larger here than in syn, yet still much lower intensity due to higher energy conformation
CCSD(T)/ANO1

16. Likely series associated with anti-isomer
Loomis Wood plot (centered on aR k-1 11 series)
This series is situated ~1267 MHz away from the initial prediction for the ground state of the anti-isomer based solely on constants from initial calculations
Here, decided to show LW plot with triangles rather than full spectrum since these low-intensity lines are difficult to see in the full spectrum. You’ll notice these transitions are clear, up until about 320 GHz, at which point the transitions drop-off in intensity at a region of the spectrum in which the sample was a bit tired (one of the last sample chamber fills).

17. Summary and future work
Fits obtained for several syn-isomer vibrational states:
In progress:
Investigate remaining syn-isomer vibrational states
Examine potential perturbations between them
Examine the inertial defect of 2ν9 + ν12 in an effort to find b-type R-branch transitions and obtain a fit
Continue searching for/assigning anti-isomer ground state transitions
Mention possible states perturbing one another and how you currently have fits for some transitions, just not well converged yet
Need to examine inertial defect for 2v9+v12 because fit currently only includes bQ-branch transitions. Examining the inertial defect will give the rotational constants corresponding to the state
Fits completed using the A-reduced Hamiltonian

18. Thanks! Prof. Robert J. McMahon Prof. R. Claude Woods Research Group:
Dr. Brian J. Esselman
Brent Amberger
Ben Haenni
Zachary Heim
Matisha Kirkconnell
Stephanie Knezz
Vanessa Orr
Cara Schwarz
Maria Zdanovskaia