A serendipitous phosphonocarboxylate complex of boron: when

Electronic Supplementary Material (ESI) for Chemical Communications This journal is © The Royal Society of Chemistry 2011

SUPPLEMENTARY DATA

A serendipitous phosphonocarboxylate complex of boron: when vessel becomes reagent

Katarzyna M. Błażewska,a,b Ralf Haiges,a Boris A. Kashemirov,a Frank H. Ebetino,c Charles E. McKenna*a

a Department of Chemistry, University of Southern California, Los Angeles, CA 90089-0744, USA, Fax: +1 310 740 0930; Tel: +1 310 740 2698; E-mail:
[email protected] b Institute of Organic Chemistry, Technical University of Łódź, Zeromskiego St. 116, 90-924 Łódź, Poland; Fax: +48 42 636 55 30; Tel: +48 42 631 31 46 c Warner Chilcott Pharmaceuticals, Mason, OH 45040, USA; Tel: +1 513 622 3630

Table of Contents

S:

General methods, reagents and materials..............................................................................................................3

HPLC and NMR detection of 5 and 6 ..................................................................................................................3

pH and temperature effects (3) .............................................................................................................................3

Vessel effects (3) ...................................................................................................................................................4

pH and temperature effects (1) .............................................................................................................................4

Vessel effects (1) ...................................................................................................................................................4

HRMS studies .......................................................................................................................................................4

Detection of 11B NMR signal ..............................................................................................................................4

Crystallization of 5 compound isolated from evaporate of 3 and X-ray crystallographic analysis ......................4

Identification of the source of boron ....................................................................................................................5

3 boron complex generation from borate..............................................................................................................5

Figure S1. HPLC trace of the “ghost” compound from 3-E1 after isolation by preparative HPLC.....................5

Figure S2. Same solution as in Figure S3, after being stored at rt, pH 7.4, for 27 h, 44 h, 72 h. .........................5

Figure S3. HPLC trace of 3 enantiomers (E1, E2) from 3 racemate sample. .......................................................6

Figure S4. 1H NMR (D2O, pH 3.6, 400 MHz) of the “ghost” compound from 3-E1 after preparative HPLC isolation as above..................................................................................................................................................6

Figure S5. 1H NMR of product from reaction of 3 with aq. sodium borate (rt). Compare with Figure S4. ........7

Figure S6. 31P NMR of product from reaction of 3 with aq. sodium borate (rt). Compare with Figure 3 in the manuscript. ............................................................................................................................................................7

Figure S7. FAB HRMS of the isolated “ghost” compound derived from 3-E2. ..................................................8

Figure S8. ORTEP drawings of both (S,S)-3 and (R,R)-3 dimer boron complexes formed in the crystal. ..........9

Figure S9. Structures of diastereomeric dimer boron complexes. ........................................................................9

Figure S10. Unit cell of the (R,R)-3 and (S,S)-3 dimer boron complex.............................................................10

Table S1. Crystal data and structure refinement for C20H14BN4O16P2. .........................................................11

Table S2. Atomic coordinates (× 104) and equivalent isotropic displacement parameters (Å2 × 103) for C20H14BN4O16P2. ...........................................................................................................................................12

Table S3. Bond lengths [Å] and angles [°] for C20H14BN4O16P2. .................................................................13

Table S4. Anisotropic displacement parameters (Å2 × 103) for C20H14BN4O16P2. ......................................14

Electronic Supplementary Material (ESI) for Chemical Communications This journal is © The Royal Society of Chemistry 2011
Table S5. Hydrogen coordinates ( × 104) and isotropic displacement parameters (Å2 × 103) for C20H14BN4O16P2. ...........................................................................................................................................15 Parameters and numerical results for Spartan calculations of dimer complexes of boron. ................................16 Reference List .....................................................................................................................................................19
S2

Electronic Supplementary Material (ESI) for Chemical Communications This journal is © The Royal Society of Chemistry 2011
General methods, reagents and materials Compound 1 was a gift from Procter & Gamble Pharmaceuticals, Inc. Compound 3 was synthesized by our
published method.1 All solvents and reagents were reagent grade, purchased commercially from Sigma-Aldrich, Inc. and used without further purification except as mentioned below. Triethylamine was freshly distilled before use. HPLC was carried out using a Dynamax Rainin Model SD – 200 pump equipped with a Shimadzu SPD – 10A VP UV Vis detector on ProntoSIL AX QN 8 × 150 mm and ProntoSIL AX QD 4 × 150 mm chiral columns (Bischoff Chromatography, Leonberg, Germany). The columns were eluted isocratically with 0.7 M TEAA in 75% MeOH at pH 5.8 for enantioseparation of 3, or with 0.25 M TEAA in 75% MeOH at pH 6.9 for enantioseparation of 1. NMR spectra were measured on Varian Mercury 400 and Bruker 250 spectrometers. Chemical shifts (δ) are reported in parts per million (ppm) relative to internal residual HDO in D2O (pH ~12, δ 4.7) (1H) or ext. H3PO4 (31P). Standard MS was performed using a Thermo-Finnigan LCQ DECA XPmax Ion Trap LC/MS/MS equipped with an ESI probe. HRMS analysis was performed at the UC Riverside MS Facility, using a VG-ZAB MS spectrometer operated in FAB mode. Mass spectra simulations were calculated using the iMass program for Apple Mac computers. UV spectra were acquired on a Beckmann Coulter DU 800 spectrophotometer. Concentrations of 3 and 5 in aqueous solution (pH 6.3) were assigned from their UV spectra, using ε = 6450 at 280 nm, as determined for 3.1
Geometric direct minimization calculations were performed using the SPARTAN '08 Quantum Mechanics software suite: Release 132v4, using the RHF-SCF method and the 3-21G* basis set.
HPLC and NMR detection of 5 and 6 In chiral HPLC separations of the 3 enantiomers, fractions were collected into “fresh” borosilicate glass test
tubes (VWR, Inc.). Reanalysis of the isolated enantiomer fraction (~1-2 mg) after concentration by evaporation often revealed the presence of a small, early-eluting extra peak (Figure 2 in the manuscript). Due to its ephemeral nature (it decreased on standing or disappeared on heating in aq. solution at pH ≥ 7), it was initially referred to informally as a “ghost” compound. The unprocessed 3 racemate did not give a detectable “ghost” peak on HPLC analysis (cf. Figure 2 in the manuscript). A similar phenomenon was observed with 1 enantiomer separations (data not shown), and it was found that enrichment in the “ghost” products 6 could be achieved by keeping sample size small (Figure 2/inset in the manuscript).
The “ghost” species from 3 (E1 enantiomer) could be isolated by preparative HPLC, giving a compound which when dissolved in 0.7 M TEAA in 75% MeOH at pH 5.8, 22 °C gave a single HPLC peak, ret. time 9.15 min at 1 mL/min elution from the AX QD column (Figure S1). After standing for 72 h at ~22 ºC, at pH 7.4, reanalysis on the AX QD column showed disappearance of this peak, with one new peak created with ret. time 12.8 min, identified as 3-E1 (Figure S2). An HPLC analysis of racemic 3 is presented in Figure S3.
The “ghost” compound 5 derived from 3-E1(or from 3-E2) has distinct 1H (Figure S4) and 31P (Figure 3 in the manuscript) NMR spectra, the latter showing a single peak at δ 11.9 in D2O at pH 3.6, upfield from 3 itself (δ ≥ 13). 31P NMR values for 3 were δ 12.9-13.0 (pH 2.8) to δ 15.6 (pH 7); for the 3 “ghost” (major) δ 11.8 − 11.9 (pH 2.8-7). This permitted convenient exploration of conditions that might affect the formation and stability of the “ghost” compounds derived from 3.
The following experiments were carried out on 0.5 − 1 mg samples of 3 on preparative HPLC isolates of the 3 enantiomers after collection into “fresh” borosilicate glass test tubes, evaporation and storage in “fresh” borosilicate glass vials, unless otherwise noted.
The 31P NMR singlet of 6 at δ 12.2 ppm (pH 4.8), was accompanied by two other small singlets at δ ppm, both with approximately the same integration. These were not investigated, but might correspond to an asymmetric dimer.
pH and temperature effects (3) a. 3-E1, 18 mM (UV) in H2O, pH 6.2, heated to 90 °C, cooled to rt, sealed in a fresh glass ampule and kept at 5 − 7 °C for 1 yr: 3, 90%; “ghost”, 10% (HPLC). b. As in a. but, after 4 mos: 3, 96%, “ghost”, 4%. c. As in a. but, pH 7.5, after 3 mos: 3, 100%. d. As in a. but, pH 7.0, after 13.5 mos: 3, 100%. e. As in a. but, pH 4.8, after 4 d: 3, 70%, “ghost”, 30%. f. As in a. but, 3 − 6 mM evaporated at pH 2.8 (H+ Dowex); then pH adjusted to 7: initially, 3 60%, “ghost”, 40%; immediately after pH 7, 3 59%, “ghost” 36%; new peak at δ 11.5 (5%). g. As in a. but, adjust pH to 6.5, 85 °C, 2 h or 4 d, rt; or 1 d, 5 °C: 3, 100%. It was concluded that acidic pH during evaporation favored “ghost” product formation, whereas near neutral
S3

Electronic Supplementary Material (ESI) for Chemical Communications This journal is © The Royal Society of Chemistry 2011
pH, it was converted back to 3; this process was accelerated by heating in water. Similar results were obtained with solutions containing 5 mM MgCl2 (data not shown).
Vessel effects (3) a. Vials rinsed with EtOH and then with H2O and dried gave somewhat reduced yields of the 5 product at pH 2.8. b. When polypropylene vials were substituted for the glass vials, no 5 product formation could be detected under any of the above conditions.
pH and temperature effects (1) a. 1, 30 mM, pH 4: 1, 87% (δ 15.0), “ghost” product, 9% (δ 12.1); 47%-49% after 20 h, 5 °C or 42%-55% after 7 d rt. b. 1, ~33 mM, pH 2.5, 20 h: precipitate. c. 1, 33 mM, pH 4.6, rt or after 20 h (5 °C): 95% 1, ~4-5% “ghost” product. d. At pH ≥ 7.4 and 5 °C, no “ghost” product was detected after up to 15 d storage.
Vessel effects (1) a. 1, 11-16 mM aq. solutions, pH 4.6 − 5.3 after evaporation in repeatedly washed (dil. HCl) borosilicate glassware and storage in polypropylene vials, or in a teflon bottle gave directly after evaporation, or after 2, 5, or 7 d storage, 100% 1 by 31P NMR (no detectable “ghost” product). It was concluded that the same pattern as with 3 was observed; however at low pH 1 gives a precipitate whose
composition was not determined.
HRMS studies The purified 5 isolate from 3-E2 was subjected to high resolution mass spectrometric analysis using a FAB
instrument. The dominant peak was observed at 579.0488 M/Z, an excellent fit for 12C201H1816O1214N431P211B: [M − H]− calcd, 579.049 but was a poor fit to a carbon-centered complex ([M − H]− calcd, 579.0318). The predicted [M − H −1]− peak at ~20% of the [M − H]− peak intensity, due to a naturally abundant 10B ion, is also observed (Figure S5).
Detection of 11B NMR signal A comparable sample was found to have a broad resonance at ~7-8 ppm (160.42 MHz quartz sample tube;
relative to ext. 15% BF3 • OEt2 in CDCl3.
Crystallization of 5 compound isolated from evaporate of 3 and X-ray crystallographic analysis Racemic 3 (2 mg) was placed in a fresh borosilicate vial (4 mL) and dissolved in 250 µL H2O (heating applied
to complete dissolution); 9 µL TFA was added. The mixture was brought to rt and the solution evaporated in the air for 10 d to give a solid crystalline residue.
The single crystal X-ray diffraction data was collected on a Bruker 3-circle platform diffractometer equipped with a SMART CCD (APEX) detector with the χ-axis fixed at 54.74° and using Mo Kα radiation (Graphite monochromator) from a fine-focus tube. The diffractometer was equipped with a Cryo Industries Cryocool-LN2 apparatus for low-temperature data collection, using controlled liquid nitrogen boil-off. The crystals were mounted on a goniometer head using a CryoLoop and PFPE oil. Cell constants were determined from 60 tensecond frames. A complete hemisphere of data was scanned on omega (0.3°) with a run time of ten seconds per frame at a detector resolution of 512 × 512 pixels using the SMART software package.2 A total of 1,271 frames were collected in three sets and a final set of 50 frames, identical to the first 50 frames, was also collected to determine any crystal decay. The frames were then processed on a PC, running Windows 2000 software, by using the SAINT software package3 to give the hkl files corrected for Lp/decay. The absorption correction was performed using the SADABS program.4 The structure was solved by the direct method using the SHELX-90 program and refined by the least squares method on F2, SHELXL-97 incorporated in SHELXTL Suite 6.12 for Windows NT/2000.5 All non-hydrogen atoms were refined anisotropically. For the anisotropic displacement parameters, the Ueq is defined as one third of the trace of the orthogonalized Uij tensor. ORTEP drawings were prepared using the ORTEP-3 for Windows V2.02 program.6 Further details of the crystal structure investigations reported in this paper may be obtained from the Cambridge Crystallographic Data Centre (CCDC), 12 Union Road, Cambridge CB2 1EZ (UK) (fax: +44 (1223)336-033 or email: [email protected]) by quoting the depository number CCDC 794952.
The crystal structure analysis revealed a mixture of (R,R) and (S,S) 3 dimer boron complexes (Figure S8-S9)
S4

Electronic Supplementary Material (ESI) for Chemical Communications This journal is © The Royal Society of Chemistry 2011
formed for the crystal. Identification of the source of boron
It was confirmed that the 5 compound was not formed from 3 under any of the following conditions: 1) exposure to the HPLC column without isolated sample evaporation; 2) prolonged exposure to TEAA buffer at pH 4-6 in repeatedly washed (dil. HCl) borosilicate glassware; 3) evaporation in repeatedly washed (dil. HCl) borosilicate glassware; 4) collection of HPLC fractions into polypropylene test tubes and evaporation in repeatedly washed (dil. HCl) borosilicate glassware. However, when HPLC fractions of 3 were collected into “fresh” VWR borosilicate glassware test tubes, a “ghost” peak was reproducibly detected. 3 boron complex generation from borate
3 (2.2 mg, 0.0077 mmol) was suspended in 0.5 mL H2O and heated to dissolve (pH 2.7). To the resulting solution was added 0.7 M H3BO3 aq. (0.5 equiv, no change in pH). After evaporation to dryness at 100 ºC (oil bath 4 h) a white solid was obtained that was suspended in small amount of D2O, and dissolved by addition of 2.25 M aq. NaOH (final pH 5.7). The conversion of 3 into the product was 89% by 31P NMR.
Figure S1. HPLC trace of the “ghost” compound from 3-E1 after isolation by preparative HPLC. Conditions: AX QD 0.7 M TEA/AcOH in 75% MeOH, pH 5.8, flow rate 1 mL/min.
E1-3 ghost peak
9.4
Figure S2. Same solution as in Figure S3, after being stored at rt, pH 7.4, for 27 h, 44 h, 72 h. Conditions: AX QD 0.7 M TEA/AcOH in 75% MeOH, pH 5.8, flow rate 1 mL/min.
S5

Electronic Supplementary Material (ESI) for Chemical Communications This journal is © The Royal Society of Chemistry 2011
Figure S3. HPLC trace of 3 enantiomers (E1, E2) from 3 racemate sample. HPLC conditions: as above, AX QD column, flow rate = 1mL/min.
Figure S4. 1H NMR (D2O, pH 3.6, 400 MHz) of the “ghost” compound from 3-E1 after preparative HPLC isolation as above.
S6

Electronic Supplementary Material (ESI) for Chemical Communications This journal is © The Royal Society of Chemistry 2011
ppm 3.70 3.60 ppm (t) 10.0
Figure S5. 1H NMR of product from reaction of 3 with aq. sodium borate (rt). Compare with Figure S4.
ppm (t) 30.0
Figure S6. 31P NMR of product from reaction of 3 with aq. sodium borate (rt). Compare with Figure 3 in the manuscript.

Electronic Supplementary Material (ESI) for Chemical Communications This journal is © The Royal Society of Chemistry 2011
Figure S7. FAB HRMS of the isolated “ghost” compound derived from 3-E2. (Lower spectrum is an enlargement of the upper spectrum; inset, predicted isotopic distribution pattern for C20H18O12N4P2B near M/Z 579 predicting a 20% relative abundance [M − H −1]− peak for a 10B ion.
S8

Electronic Supplementary Material (ESI) for Chemical Communications This journal is © The Royal Society of Chemistry 2011
Figure S8. ORTEP drawings of both (S,S)-3 and (R,R)-3 dimer boron complexes formed in the crystal. Thermal ellipsoids are shown at the 50% probability level.
Figure S9. Structures of diastereomeric dimer boron complexes. The structure calculated by geometric direct minimization using the Spartan ‘08 Quantum Mechanics software suite. Left: heterochiral, (R,S) dimer complex. Right: homochiral, (R,R) dimer complex (cf. Figure S8, S10). Thermal ellipsoids are shown at the 50% probability level.
S9

Electronic Supplementary Material (ESI) for Chemical Communications This journal is © The Royal Society of Chemistry 2011
Figure S10. Unit cell of the (R,R)-3 and (S,S)-3 dimer boron complex. Thermal ellipsoids are shown at the 50% probability level.
S10