Dendritic polarizing agents for DNP SENS† †Electronic supplementary information (ESI) available. See DOI: 10.1039/c6sc03139k Click here for additional data file.

Dendrimer-shielded polarizing agents for the application of DNP SENS to reactive surfaces.

. Diffusion coefficients and hydrodynamic diameters of the dendrimers S3 Figure S1. Structures of the corresponding non-dendritic PAs S3 II. Electronic paramagnetic resonance spectroscopy S4 a. Continuous wave (CW) EPR S4 Figure S2. X band CW EPR spectra S5 b. Pulse EPR spectroscopy S5 Figure S3. Inversion recovery pulse sequence in pulse EPR to determine T1e Figure S4. T1e relaxation traces (blue) and fits (red) S5 S6 Figure S5. Hahn echo pulse sequence in pulse EPR to determine T2e S8 S9 c. DNP performances of dendritic biradiclas on materials S9 d. DNP properties of bTUreaB series on P@SiO2 S10 e. DNP on W13C@SiO2 S11 Figure S8. DNP enhanced 29 Si CP-CPMG spectra S12 Figure S9. DNP enhanced 13 C CP spectrum of W13C@SiO2 S13 Figure S10. DNP enhanced 13 C CP spectrum of W13C@SiO2 S14 IV. Experimental section of synthesis S15 a. General description S15 b. Synthesis and characterization S16 Table S2. Experimental details of all ester bond coupling reactions S16 Table S3. Experimental details of all amide bond coupling reactions S18

I. Hydrodynamic diameters of the dendrimers
The hydrodynamic diameters of the dendrimers were estimated by measuring their diffusion coefficients in dilute CDCl3 solution with diffusion-ordered spectroscopy (DOSY). Measurements were done on a Bruker 400MHz NMR spectrometer with a standard 5 mm probe for solution samples. The temperature of the sample was kept at 25 o C to avoid solution convection. 64 increments were collected for each experiment. The intensity decay curves were optimized to be within 95% and 5% region depending on the 1 H T1 relaxation, and were fit with the topspin T1/T2 fitting module to afford the corresponding diffusion coefficients. Hydrodynamic diameters were calculated based on the Stokes-Einstein equation.

II. Electronic paramagnetic resonance spectroscopy a. Continuous wave (CW) EPR
Continuous Wave (CW) EPR spectra were recorded on a Bruker EMX X Band spectrometer (9.5 GHz microwave frequency). The conversion time was set to 40.96 ms and the time constant to 5.12 ms. 1024 data points were recorded. The modulation frequency was 100 kHz and the modulation amplitude was 1 G. In all measurements, attenuation was varied such that no saturation was observed.

Spin count
The samples were filled in Hirschmann glass capillaries, which were then closed with putty. The sample position in the cavity was carefully optimized. The spectra were recorded at room temperature with a sweep width of 600 G and an attenuation of 26 dB. The amount of radical was determined by double integration of the CW spectra and by referencing to a calibration curve of TEMPO solutions in TCE measured for the concentration range between 0.4 and 80 mM. Data was processed with MATLAB ® (R2011a, MathWorks Inc.).

Peak-to-peak line width
Each sample was prepared as a 30 mM TCE solution and filled in a 3.0 mm quartz tube. All spectra were recorded at 110 K using a nitrogen flow cryostat. Attenuation was 40 dB. The EPR spectrum of a nitroxide radical consists of three lines due to strong hyperfine interaction with the 14 N nucleus. For the line width measurements we have used the central line, which is the least broadened by the g-tensor and hyperfine anisotropies and which therefore is the most sensitive to the dipolar broadening. For the obtained signal-to-noise ratios, the estimated line width errors were within 5 %.

b. Pulse EPR spectroscopy
All pulse experiments were recorded at W band (94 GHz) on a Bruker Elexsys E680 EPR spectrometer. The temperature was stabilized with an Oxford helium flow cryostat and was held at 100 K for all experiments. All samples were prepared as 16 mM TCE solutions to be close to actual DNP conditions.

T1e Measurements
The longitudinal relaxation time T1e was determined by performing an inversion recovery experiment. The used pulse sequence is shown in Figure S3. Pulse lengths used were 80 ns for the inversion pulse, 40 ns for the π/2 and 80 ns for the refocusing π pulse, respectively. The initial dvar was 1000 ns, and the time increment was 500 ns. Error bars are estimated to be approximately 5 % as good signal-to-noise traces were recorded for all the samples. Figure S3. Inversion recovery pulse sequence in pulse EPR to determine T1e 40ns 80ns

S6
The inversion recovery time traces were fit using a stretched exponential function (see below) and the first moment of the distribution is discussed in the text as a mean relaxation time.
Stretched exponential function was used for fitting of the inversion recovery traces on top and the function to determine the mean relaxation time <T1e> on the bottom. I0 is the initial intensity, I1 the proportionality factor, T1e * the decay time parameter and β the stretching parameter. The mean relaxation time <T1e> discussed in the text is the first moment of the distribution and is determined using the decay time parameter T1e * and the stretching parameter β.

T2e Measurements
The transverse relaxation time T2e was determined using a variable-delay Hahn echo pulse sequence. The used pulse sequence is shown in Figure S4. Pulse lengths of 40 ns for the π/2 and 80 ns for the refocusing π pulse were used. The initial dvar was at 400 ns, the time increment was 4 ns. The two-pulse echo decay traces were fitted using a monoexponential function. Error bars are estimated to be approximately 5 % as good signal-to-noise traces were recorded for all the samples.

III. DNP enhanced NMR spectroscopy a. General description
All DNP enhanced solid-state NMR experiments were either conducted on a Bruker 600 MHz (14.1 T) spectrometer using a 3.2 mm HX or HXY probe located at ETH Zürich, or on a Bruker 400 MHz (9.4 T) spectrometer using a 3.2 mm HXY probe located at the Centre de RMN à Trés Hauts Champs at ISA Lyon. All samples were cooled to 100 K by a cryogenic heat exchanger system. Microwaves used to drive the DNP Cross Effect are provided by gyrotrons emitting at 263 GHz (400 MHz spectrometer) and 395 GHz (600 MHz spectrometer) with power between 6 to 10 W. The magnetic fields were adjusted by referencing the higher frequency peak of adamantane to 38.5 ppm. TCE peak was used as a second reference for 13 C spectra. For moisture or air sensitive materials, samples were prepared in an argon-filled glove box and inserted in the cryogenic probe within a short period of time (~ 5 min). The dendritic PAs and TEKPol used were dried under high vacuum (10 -4 mbar) for 16 hours and 5 hours, respectively. The solvent used, 1,1,2,2-tetrachloroethane (TCE), was dried over calcium hydride overnight, then vacuum transferred to a rotaflo flask. All materials were stored in the argon-filled glove box after removing residual moistures and degassing.

b. Bulk solution enhancement
The monoradical was prepared as a 30 mM TCE solution; the biradicals were prepared as 16 mM solutions. 23 μL of the prepared solution were pipetted into a 3.2 mm sapphire rotor, which was then sealed with a siloxane insert to avoid solution leaking. Each sample was deoxygenated by a series of freeze-thaw cycles, i.e. insert-eject cycles, inside the cryogenic DNP probe to reach the maximum DNP enhancement.
Ramped cross polarization (CP) from 1 H to 13 C was used for all experiments with contact time between 0.5 to 2.0 ms. SPINAL64 2 was used for 1 H decoupling at γB1 of 100 kHz. The enhancements were measured by comparing the intensity of the 13 C spectrum when microwaves were on to the intensity of the 13 C spectrum when microwaves were off. The 1 H T1 was measured by saturation recovery experiments. All experiments were carried out with MAS frequency set to 8kHz.

c. DNP performances of dendritic biradiclas on materials
All samples were measured on a Bruker 600 MHz (14.1 T) DNP spectrometer. Ο v * spinning side bands Ο polysiloxane plug Δ dendrimer siignal v pentachloroethane S10 31 P was used for all experiments with a contact time of 1.5 ms. SPINAL64 2 was used for 1 H decoupling at γB1 of 100 kHz. 1 H T1 was measured by saturation-recovery experiment, and the recycle delay was set to be 5 x T1 as 30 s (for the experiments described in SI III.d). The enhancement was measured by comparing the intensity of the 31 P spectrum when microwaves were on to the intensity of the 31 P spectrum when microwaves were off. 10 wt.-% Sn-beta zeolite was incipient wetness impregnated with 16 mM PyPolB-D2[G3] TCE solution. 1 H-119 Sn CP-CPMG experiments were performed to acquire the spikelet 119 Sn NMR spectra with the MAS frequency set to 10 kHz. 20 echoes were collected for each experiment. Ramp CP from 1 H to 119 Sn was used for all experiments with a contact time of 1.0 ms. SPINAL64 2 was used for 1 H decoupling at γB1 of 100 kHz. 1 H T1 was measured by saturation-recovery experiment, and the recycle delay was set to be 1.3 x T1 as 3.2 s. The enhancement was measured by comparing the intensity of the 119 Sn spectrum when microwaves were on to the intensity of the 119 Sn spectrum when microwaves were off.

P@SiO2
TBP (moisture sensitive) was incipient wetness impregnated with 16 mM PyPolB-D2[G3] TCE solution. 1 H-31 C CP experiments were performed to acquire the 13 C NMR spectra with the MAS frequency set to 10 kHz. Ramp CP from 1 H to 31 C was used for all experiments with a contact time of 2.0 ms. SPINAL64 2 was used for 1 H decoupling at γB1 of 100 kHz. 1 H T1 was measured by saturation-recovery experiment, and the recycle delay was set to be 1.3 x T1 as 4.5 s. The enhancement was measured by comparing the intensity of the 13 C spectrum when microwaves were on to the intensity of the 13 C spectrum when microwaves were off.

d. DNP properties of bTUreaB series on P@SiO2
P@SiO2 (moisture sensitive) was separately impregnated with 16 mM bTUreaB-PAs TCE solutions. Each sample was prepared with the ratio of 1 mg of solid to 1μL of 16mM biradical solution, and the exact masses used were recorded. All measurements were done on a Bruker 600 MHz DNP spectrometer with the MAS frequency set to be 10 kHz. 1 H-31 P CP Hahn-echo experiments were performed to acquire the 31 P NMR spectra. Ramp CP from 1 H to 31 P was used for all experiments with a contact time of 1.5 ms. SPINAL64 2 was used for 1 H decoupling at γB1 of 100 kHz. 1 H T1 was measured by saturation-recovery experiment. 31 P T2 was measured by varied-delay CP Hahn-echo experiments. The enhancement was measured by comparing the intensity of the 31 P spectrum when microwaves were on to the intensity of the 31 P spectrum when microwaves were off. The contribution factor measurements were S11 adjusted from the procedure described previously. 3 The signal integral of the sample impregnated with the biradical solution with microwave off was compared to the signal integral of the sample with pure TCE impregnation. The integral was normalized by mass and by number of scans. The recycle delay was set to be 5 x T1 in order to complete full relaxation. Accordingly the value of 1-θ represents the fraction that does not contribute to the observable signal.

e. DNP on W13C@SiO2
W13C@SiO2 (moisture and air sensitive) was separately impregnated with 16 mM PyPolB-D2[G3] and TEKPol TCE solutions. All measurements were done on a Bruker 400 MHz DNP spectrometer. The MAS frequency was set to be 12 kHz or 10 kHz for 1D experiments and 5 kHz for the CPMAT experiment. Ramp CP from 1 H to 13 C was used for all 13 C CP experiments with a contact time of 0.5 ms. Ramp CP from 1 H to 29 Si was used for all 29 Si CP-CPMG experiments with a contact time of 7.0 ms, and 30 echoes were collected for each experiment. SPINAL64 2 was used for 1 H decoupling at γB1 of 100 kHz. 1 H T1 was measured by a saturation-recovery experiment, and the recycle delay was set to be 1.3 x T1. The enhancement was measured by comparing the intensity of the spectrum ( 13 C or 29 Si) when microwaves were on to the intensity of the spectrum ( 13 C or 29 Si) when microwaves were off.

IV. Experimental section of synthesis a. General information
For reactions carried out under N2, a standard N2-filled Schlenk line was used.
Other reactions were performed in atmospheric environment. Dry dichloromethane (DCM), diethyl ether (Et2O) and toluene were purified by passage through double solvent purification alumina columns (MBraun). Dry tetrahydrofuran (THF) was dried over sodium with benzophenone as indicator, and further vacuum transferred and stored in a rotaflo flask. Dry dimethylformamide (DMF) was purchased from Sigma-Aldrich (99.8% extra dry over Molecular Sieve) and used as received. Other solvents: Pentane, cyclohexane, Et2O and ethyl acetate (EA) were purchased from the ETH HCIshop Tanklager and used as received. DCM and THF were purchased from Acros and used as received. Solution NMR measurements were done on Bruker DPX-300 MHz or 400 MHz or 500 MHz NMR spectrometers (specified in the following text) using CDCl3 as a solvent. The 1 H and 13 C chemical shifts are referenced to the chloroform peak (7.26 ppm for 1 H and 77.00 ppm for 13 C). High-resolution mass spectra were measured by the MS-service of the Laboratorium für Organische Chemie, ETH Zürich and by Aix-Marseille Université Mass Spectrum Facility, Spectropole Saint Jérôme Marseille. Values are given as m/z.
The chloroplatinic acid hexahydrate DCM solution (1 equiv.) was mixed with the tetrabutylammonium hydrogensulfate water solution (2 equiv.) for 3 hours. The organic layer was separated from the aqueous layer and dried over magnesium sulfate. The solvent was removed in vacuo and afforded an orange solid. The orange solid was used as bis(tetrabutylammonium) hexachloroplatinate (Lukevics' catalyst) in the following reaction directly without further purification.

b. Synthesis and characterization Ester bond coupling
The ester coupling reagent (DCC or EDCI HCl salt, 1.1 equiv., Table S4) and DMAP (catalytic amount, approximately 10 mol%) were added into a solution of a OHfunctionalized substrate (D1[G3] or D1[G4], 1 equiv., or propyl alcohol, 2.5 equiv., the amount of material used, WOH is specified in Table S4.) and a COOH-functionalized substrate (4-carboxy TEMPO or bTUreaB, 1 equiv., the amount of material used, WCOOH is specified in Table S4.) in dichloromethane (2 mL). After the mixture was stirred for 16 hours at room temperature, it was poured into water and extracted with dichloromethane (50 mL x 3). The organic layers were collected, washed with water (50 mL x 3) and brine (50 mL) and dried over sodium sulfate. The solvent was removed in vacuo. The residue was purified by preparative thin layer chromatography, and the corresponding eluent was cyclohexane/EA (15/1 for TEMPO-PAs, and 2/1 for bTUreaB-PAs). The final product was extracted by stirring the product-adsorbed silica gel in EA (100 mL). After stirring for 10 min at room temperature, the slurry was filtered, and the solvent was removed in vacuo. The final residue was re-dissolved in dichloromethane (less than 5 mL) and filtered through a neutral cellulose syringe filter to remove residual silica gel. The solvent was removed in vacuo to afford the final product (red solids for TEMPO-C3 and bTUreaB-C3; red sticky oils for TEMPO-

Amide bond coupling
The amide coupling reagent (EDCI HCl salt, 1.5 equiv., Table S5), HOBt (catalytic amount, approximately 10 mol%) and diisopropylamine (1 equiv., Table S5) were added into a solution of D2[G3] (1.1 equiv) or propyl amine (3 equiv.) (the amount of material used, WNH2 is specified in Table S5) and a COOH-functionalized substrate (4carboxy TEMPO or bTUreaB or PyPolB (1 equiv.) (the amount of material used, WCOOH is specified in Table S5) in dichloromethane (2 mL). After the mixture was stirred for 16 hours at room temperature, it was poured into water and extracted with dichloromethane (50 mL x 3). The organic layers were collected, washed with water (50 mL x 3) and brine (50 mL) and dried over sodium sulfate. The solvent was removed in vacuo. The residue was purified by preparative thin layer chromatography, using pentane/EA (1/5 for TEMPO-D2[G3] and 1/2 for bTUreaB-D2[G3]), pure EA for PyPolB-D2[G3], or DCM/methanol (10/1) for PyPolB-C3 as eluents. The final product was extracted by stirring the product-adsorbed silica gel in EA (100 mL). After stirring for 10 min at room temperature, the slurry was filtered, and the solvent was removed in vacuo. The final residue was re-dissolved in dichloromethane (less than 5 mL) and filtered through a neutral cellulose syringe filter to remove residual silica gel. The solvent was removed in vacuo to afford the final product (red sticky oils for dendritic PAs; red solid for PyPolB-C3). To a solution of ketone 2' (100mg, 0.39 mmol) and ethyl 4-aminobenzoate (70mg, 0.39 mmol) in dry THF (4 mL) were added Na2SO4 and AcOH to adjust pH at 6-7 and the solution was stirred at room temperature for 2 h. Then, Na(OAc)3BH (150mg, 0.71 mmol) was added portionwise and the reaction was stirred at 25°C for 6 h. The mixture was concentrated under reduced pressure. The residue was solubilized in CH2Cl2, washed with a saturated aqueous solution of K2CO3 35% (10mL), dried on Na2SO4, concentrated under reduced pressure and the crude product was purified by SiO2 column chromatography with CH2Cl2/EtOH (9/1) as eluent to give the desired

S25
Under N2, to a solution of D1[G1](allyl) (2.02 g, 6.00 mmol) and dichlorophenylsilane (2.34 g, 13.2 mmol) in dry toluene (15.0 mL) was added platinum(0)-1, 3-divinyl-1, 1, 3, 3-tetra-methyldisiloxane complex solution (120 μL, Pt ~ 2% in xylene). The mixture was stirred for 16 h at 60 °C, then toluene and excess dichlorophenylsilane were removed in vacuo. Dry THF (30.0 mL) was then added to the residue under N2, and allylmagnesium chloride solution (13.2 mL, 26.4 mmol, 2 M in THF) was added into the mixture at 0 °C via syringe pump at a rate of 1 mL/min. The mixture was warmed up to room temperature and further stirred for 16 h at 50 °C. After cooling to room temperature, the mixture was poured into cold saturated ammonium chloride aqueous solution and extracted with pentane. The organic layer was collected, washed with water and brine, and dried over magnesium sulfate. After evaporation of solvent in vacuo, the residue was chromatographed over silica gel (cyclohexane/Et2O = 50/1) to afford

S29
To a solution of D1[G3]OBn (340 mg, 0.270 mmol) in dichloromethane (4.00 mL) was added 2, 3-dichloro-5, 6-dicyano-1, 4-benzoquinone (90.0 mg, 0.400 mmol). The mixture was refluxed for 16 h, then it was cooled to room temperature and poured into cold saturated sodium bicarbonate aqueous solution. The mixture was extracted with dichloromethane, and the organic layer was collected, washed with water and brine, and dried over magnesium sulfate. After evaporation of the solvent in vacuo, the residue was chromatographed over silica gel (cyclohexane/EA = 15/1) to afford D1[G3] as colorless sticky oil (210 mg, 66% .00 mmol) and dichlorophenylsilane (3.12 g, 17.6 mmol) in dry toluene (20 mL) was added platinum(0)-1, 3-divinyl-1, 1, 3, 3-tetra-methyldisiloxane complex solution (80 μL, Pt ~ 2% in xylene). The mixture was stirred for 16 h at 60 °C, then toluene and excess dichlorophenylsilane were removed in vacuo. Dry THF (40 mL) was then added to the residue under N2, and allyl magnesium bromide solution (17.6 mL, 35.2 mmol, 2 M in THF) was added into the mixture at 0 °C via syringe pump at a rate of 1 mL/min. The mixture was warmed up to room temperature and further stirred for 16 h at 50 °C. After cooling to room temperature, the mixture was poured into cold saturated ammonium chloride aqueous solution and extracted with pentane. The organic layer was collected, washed with water and brine, and dried over magnesium sulfate. After evaporation of solvent in vacuo, the residue was chromatographed over silica gel (cyclohexane/Et2O = 50/1) to afford D1[G3](allyl) as colorless sticky oil (5.10 g, 87%

S39
Under N2, bis(tetrabutylammonium) hexachloroplatinate (44.0 mg, 0.050 mmol) was added to a solution of D2[G1](allyl) (2.29 g, 10.0 mmol) in a mixture of dry diethyl ether (10 mL) and dry dichloromethane (20 mL). The mixture was stirred for 30 min, then trichlorosilane (4.47 g, 33.0 mmol) was added. The flask was equipped with a calcium chloride tube, and the nitrogen inlet was closed. The mixture was stirred vigorously until 1 H NMR suggested full conversion of all allyl end groups, usually after 1 to 2 days reaction time. Solvents and excess trichlorosilane were removed in vacuo, and dry diethyl ether (40mL) was introduced to dissolve the residue. To this mixture was slowly added allylmagnesium chloride solution (49.5 mL, 99.0 mmol, 2 M in THF) via a syringe pump within 120 min at 0 °C. The mixture was warmed to room temperature. After stirring for 4 hours, it was poured into cold saturated ammonium chloride aqueous solution. The mixture was twice extracted with diethyl ether, and the organic layers were collected, washed with water and brine and dried over magnesium sulfate. After evaporation of solvent in vacuo, the residue was flash chromatographed over silica gel (pentane to pentane/diethyl ether = 50/1 to pentane/diethyl ether = 20/1) to afford D2[G2](allyl) as colorless oil (5.87 g, 86%

S41
Under N2, bis(tetrabutylammonium) hexachloroplatinate (41.0 mg, 0.046 mmol) was added to a solution of D2[G2](allyl) (3.14 g, 4.58 mmol) in a mixture of dry diethyl ether (20 mL) and dry dichloromethane (40 mL). The mixture was stirred for 30 min, then trichlorosilane (6.14 g, 45.3 mmol) was added. The flask was equipped with a calcium chloride tube, and the nitrogen inlet was closed. The mixture was stirred vigorously until 1 H NMR suggested ful conversion of all allyl end groups, usually after 3 to 4 days reaction time. Solvents and excess trichlorosilane were removed in vacuo, and diethyl ether (150mL) was introduced to dissolve the residue. To this mixture was slowly added benzylmagnesium chloride solution (64.0 mL, 128 mmol, 2 M in THF) via a syringe pump within 150 min at 0 °C. The mixture was warmed to room temperature. After stirring for 5 hours, it was poured into cold saturated ammonium chloride aqueous solution. The mixture was three times extracted with diethyl ether, and the organic layers were collected, washed with water and brine and dried over magnesium sulfate. After evaporation of solvent in vacuo, the residue was flash chromatographed over silica gel (pentane/diethylether/dichloromethane = 18/1/1 then 8/1/1) to afford