1. An Improved CP-FTMW Analysis of the Structures of Phenol Dimer and Trimer Nathan A. Seifert, Cristobal Perez, Amanda L. Steber, Daniel P. Zaleski, Justin L. Neill, Brooks H. Pate University of Virginia Alberto Lesarri Universidad de Valladolid

2. Previous Experimental Structural Studies: CP-FTMW; previously by us: RC11, 2011 (Steber et al.) 1 Substitution structure assigned, issues with Kraitchman agreement Less assigned transitions  poorer determination of constants, distortion  poorer structure Rotational coherence spectroscopy 2 and UV-Vis 3 ; Some issues with these results: O ⋯ O lengths not determined (RCS) or anomalously long (UV-Vis ), ≥3 Å Poor precision with respect to FTMW Phenol trimer detected using UV/IR double resonance, but no structural information 4 IR data consistent with symmetric “barrel” structure 1.A. L. Steber, et al. Faraday Discuss., 2011, 150, 227. 2.M. Schmitt, et al., ChemPhysChem, 2006, 7, 1241. 3.A. Weichert, et al., J. Phys. Chem. A, 2001, 105, 5679. 4.T. Ebata, et al. J. Phys. Chem., 1995, 99, 5761. Inter-ring dispersion motif: π stacking-like, or C-H ⋯ π dominated? Unsolved Mysteries of (phenol) 2 : (H2O) 2 : r[O ⋯ O] is 2.98 Å; is hydrogen bonding a major player in (phenol) 2 ? Necessitated reduced bandwidth measurement to assign isotopologues in natural abundance On a side note: Introduction - Experiment UV/Vis r 0 structure M06-2X/6-311++g(d,p)

3. Introduction - Theory (phenol) 2 : Important benchmark molecule for modeling the interplay of various non-covalent interactions ---- H-bonding: Electrostatic dominated ------------------- ------------- π-π: Dispersion (exchange) dominated CH-π: dispersion/ electrostatic competition Important to fine tune the computational balance between electrostatics and exchange for biochemical or (in this case) spectroscopic predictions

4. Experimental New horns enable use of 5 pulsed nozzles at 2-8 GHz (previous setup was 2 nozzles) New TWT with 600W peak output (previous was 200W) New oscilloscope (Tek DPO73304D) enables detection of 8 frames at 3.3 Hz! (previous 2-8GHz rate was ~0.5Hz, at 4 frames/acq!) 9.2 million averages, 2-8 GHz Dynamic range: 30000:1 4000 lines over 4:1 SNR See Cristobal’s talk, TH10, for more information on the instrument At 3.3 Hz, this is about 4 days of averaging!

5. V(θ) = (0.232 cm -1 )θ 2 - (0.00447 cm -1 )θ 3 Δ = ±4° (1σ) MO6-2X/6-311++g(d,p) scan of the hinge potential Calculating ψ with this potential gives a QM hinge fluctuation of Ab initio PES fits with R 2 = 0.99 to: Fundamental mode is 28 cm -1 by these calculations; experimentally 1 it’s 9 cm -1 Smaller than the σ seen in the theoretical results (next slide)! Phenol Dimer - Analysis Question: Will this be a problem for our determined structure? “Hinge potential” – anharmonic and large amplitude? 1. M. Schmitt, et al. J. Chem. Phys., 1995, 103, 9918.

6. Answer: Yes and no. Experimental determination is quite acceptable, but there are issues with theory. Phenol Dimer - Analysis A (MHz)B (MHz)C (MHz)Hinge Angle MP2/6-311g(d,p) 1165.2 410.3363.349.0 MP2/6-311++g(d,p) 1365.2 325.8305.769.5 MP2/cc-pVTZ-cp a 1286.0356.1317.6 64.4 CCSD/6-311g(d,p) 1545.2282.4275.6 88.7 CCSD/6-311++g(d,p) 1459.6308.9286.8 79.5 M06-2X/6-311g(d,p) 1352.7 342.1318.368.2 M06-2X/6-311++g(d,p) 1382.5 336.6306.459.6 M06-2X/6-311++g(df,pd) 1382.9 338.4307.659.6 B3LYP/6-31G(d,p) 1800.0 250.8243.7111.9 B3LYP/6-311++g(d,p) 1946.3 231.8229.6110.5 RI-DFT-D/aQz' b 1399.7 318.5292.561 CP-FTMW (r 0, this study)1415.3275(11)313.36802(32)287.96128(30)62.3(14) UV-VIS (r 0, Schmitt et al.)1416.99(39)313.51(1)288.11(1)63.0 Observations: MP2/CCSD overestimates the dispersive contributions (a known problem) B3LYP fails completely Dispersion-corrected hybrid functionals such as M06-2X (Truhlar) and RI-DFT-D (Hobza) perform the best a. P. Jurečka, J. Šponer, J. Černy and P. Hobza, Phys. Chem. Chem. Phys. 2006, 8, 1985. b. M. Koláŕ and P. Hobza, J. Phys. Chem. A, 2007, 111, 5851. A bold claim: On the basis of efficient spectroscopic discovery, new hybrid functionals might be the future.

7. Phenol Dimer - Structure Six parameters required to fit intermolecular structure: 1)R[1d ⋯ 1a] (hydrogen bond length) 2)θ[1d ⋯ 1a-2a] & θ[1a ⋯ 1d-2d] (planarity of H-bond) 3)φ[1a ⋯ 1d-2d-3d] & φ[1d ⋯ 1a-2a-3a] (“ring tilt”) 4)φ[2d-1d ⋯ 1a-2a] (hinge angle) Fix phenol monomer geometry to ab initio geometry (excellent agreement with previous expt. results) Schema (also used in UV/Vis study): r 0 intermolecular fit structure: Our “toolkit”: All 13 C and both 18 O isotopologues in natural abundance 10 3 better certainty on parent species constants (plus distortion!) UV/Vis “toolkit”: C 6 H 5 -OD measurements 13 C at the 1- position Just A/B/C on parent species with ~50-100 kHz precision

8. M06-2x/6-311++g(d,p) [frame] r 0 [spheres] Phenol Dimer - Structure M06-2x/6-311++g(d,p) [frame] r s [spheres]

9. ParameterValue (r 0 fit, this study) Value (r 0 fit, UV/VIS 1 ) M06-2X/6- 311++g(d,p) r(H d ⋯ O a ) 1.873(22)2.354(49)1.890 r(O d ⋯ O a ) 2.833(21)3.211(25)2.844 θ(O d -H d ⋯ O a ) 170.5(21)150.6(18)168.1 θ(C a -O a ⋯ H d ) 122.5(10)138.6(15)118.8 φ(O a -H a ⋯ O d -C d ) 75.5(59)109.6(45)76.8 φ(C d -O d -H d ⋯ O a ) -27.7(47)-26.5(46)-23.4 φ(C2 a -C a -O a ⋯ H d ) 10.6(17)-1.0(19)14.2 φ(C d -O d ⋯ O a -C a ) [hinge] 64.0(13)63.059.6 CdCd OdOd OaOa CaCa HdHd C2 a Take home points: Water dimer-like hydrogen bonding: (H2O) 2 : r(O ⋯ O) = 2.98(1) Å π stacking dominant, not the “twisted” C-H ⋯ π-like motif seen in UV-VIS structure Phenol Dimer - Structure Brown: UV-Vis r 0 Purple: CP-FTMW r 0 1. M. Schmitt, M. Böhm, C. Ratzer, D. Krügler, K. Kleinermanns, I. Kalkman, G. Berden and W. L. Meerts, ChemPhysChem, 2006, 7, 1241. (donor rings overlapping) CP-FTMW r 0 σ fit = 0.059 amu Å 2

10. Phenol Trimer Oblate symmetric top Parent1- 13 C6- 13 C2- 13 C5- 13 C3- 13 C4- 13 C A(MHz)--282.19749(14)282.21553(17)282.26789(23)282.23405(52)281.96181(31)281.89683(66) B (MHz)282.280790(56)281.30995(14)280.33503(17)281.44815(23)279.47608(53)280.56222(30)279.57290(65) C (MHz)--187.33(14)187.069(50)187.36(21)186.897(44)187.34(14)186.824(66) D J (kHz)0.10300(19)[0.0315] 0.0297(10)[0.0315]0.0334(12) D JK -0.13600(28)[0.0752] 0.0779(26)[0.0752]0.0725(34) DKDK --[0] dJdJ --[0.03249] 0.03227(56)[0.03249]0.03270(65) dKdK --[0] N2063875310853100 σ (kHz) 1.375.967.119.109.339.498.62 13 C isotopologues are off-symmetry axis, so they are standard oblate asymmetric tops To ease the fitting process, isotopologues were fit as pure c-type prolate asymmetric tops 1. T. Ebata, T. Watanabe, N. Mikami. J. Phys. Chem. 1995, 99, 5761. Nature of Kraitchman substitution for this system requires some assumptions regarding C-C bond lengths in the phenol monomers (ask me about it after if you’re curious)

11. Use of dummy coordinates and forced C 3v symmetry: Only three fit parameters: r, θ, φ (w/ three-fold degeneracy) Schema: Phenol Trimer - Structure Use r m (1) model to fit intermolecular structure Assume phenol monomer ab initio structure Similar assumptions to r s determination (thanks to Z. Kisiel for the helpful tips!)

12. ParameterM06-2X/6-311++g(d,p)r m (1) r(H ⋯ O) / Å1.9401.895(86) r(O ⋯ O)2.8112.760(70) r(C1 - C1')4.0103.967(83) <(O - O' - O'') / °60.060.03(73) <(O - H ⋯ O')147.3147.1(16) <(C1 - O'' ⋯ O')-23.4-27.9(14) <(C1'' - O'' ⋯ O)117.9114.0(16) <(C1 ⋯ C1' ⋯ C1'')60.060.00(75) t(C1-O-O'-C1')-6.42-5.9(28) t(O-O'-C1'-C6')85.285.2(33) c aa = c bb / amu Å 2 --1.140(85) r(O ⋯ O) shorter for phenol trimer than dimer 2.833(21) Å  2.760(70) Å Similar trend to (H2O) 2  (H2O) 3 2.98 Å  2.85 Å Stabilizing multi-body effects are apparent: “hinge” is closed: C-H/π interactions much more important 2- hydrogen with C5-C6 bond: ~2.6 Å (very typical for C-H/π) Ring planes are nearly perpendicular  85.2(33)° Phenol Trimer - Structure Theory is much more well behaved! A (MHz)B (MHz)C (MHz) M06-2X/6-311g(d,p) 199.494292.600-- M06-2X/6-311++g(d,p) 198.084290.540-- MP2/6-311g(d,p) 205.635296.244-- MP2/6-311++g(d,p) 211.380300.117-- B3LYP/6-311++g(d,p) 240.6222.786130.919 Experiment[188]282.280790(56)

13. Acknowledgements Thanks to the NSF for funding: MRI-R2, Award CHE-0960074 Pate Group Brooks Pate Cristobal Perez Simon Lobsiger Luca Evangelisti Brent Harris Amanda Steber Nathan Seifert Daniel Zaleski Newcastle University Brightspec Justin Neill Universidad de Valladolid Alberto Lesarri Thanks for your time!

14. Phenol Trimer - Structure Off-axis substitution has easy form of Kraitchman’s equations: Problem: In the above frame, |a| = 0. In PA frame, |a| ≠ 0. How do we determine rotation matrix to get PA coordinates? Solution: Assume ab-initio phenol monomer C-C bond lengths to solve Law of Cosines for required θ to convert KRA->PA (oblate basis )

15. Phenol Dimer Detected all 12 13 C and both 18 O isotopologues M06-2X/6-311++g(d,p) CP-FTMW (2-8 GHz, new) CP-FTMW (7-9 GHz, old) A (MHz) 1382.51415.32747(14) 1415.32633(48) B 336.59313.368020(41) 313.367093(60) C 306.41287.961282(38) 287.960317(66) Δ J (kHz) 0.9080.372930(76) 0.372342(65) Δ JK -5.58-3.92349(48) -3.91817(96) ΔKΔK 15.613.0664(27) 13.079(16) δJδJ 0.01820.057377(17) 0.057355(29) δKδK 0.4100.6928(19) 0.7024(85) N --481302 σ (kHz) --6.3911.1 Donor A (MHz) BC Nσ (kHz) 1- 13 C1413.0983(14)312.52409(32)287.16539(32)1208.08 2- 13 C1412.74240(48)312.75594(19)287.52814(20)1365.77 3- 13 C1403.8469(63)312.00616(35)286.73261(34)1118.25 4- 13 C1406.0358(10)310.68675(28)285.40748(28)1297.98 5- 13 C1413.37840(97)310.05088(26)285.13627(28)1338.75 6- 13 C1411.85779(88)311.08467(29)286.08891(29)1136.98 1- 18 O1387.525(25)312.54360(80)286.32269(76)6010.4 Acceptor 1- 13 C1413.23440(69)312.50304(16)287.15363(16)1235.53 2- 13 C1412.24751(35)312.72811(22)287.52476(22)1335.38 3- 13 C1403.92180(82)311.90897(22)286.72685(23)1297.11 4- 13 C1405.66540(96)310.71737(19)285.40780(19)1236.79 5- 13 C1413.02009(74)310.10090(26)285.17718(26)1216.41 6- 13 C1411.56899(64)311.12121(17)286.14677(18)1366.84 2- 18 O1388.945(42)312.70639(67)286.47370(65)697.02