Mesoporous Silica-Supported Amidozirconium-Catalyzed Carbonyl Hydroboration

Research Article

pubs.acs.org/acscatalysis

Mesoporous Silica-Supported Amidozirconium-Catalyzed Carbonyl

Hydroboration

Naresh Eedugurala, Zhuoran Wang, Umesh Chaudhary, Nicholas Nelson, Kapil Kandel,^‡

Takeshi Kobayashi, Igor I. Slowing, Marek Pruski,* and Aaron D. Sadow*

U.S. Department of Energy Ames Laboratory and Department of Chemistry, Iowa State University, 1605 Gilman Hall, Ames, Iowa

50011, United States

S

* Supporting Information

ABSTRACT: The hydroboration of aldehydes and ketones

using a silica-supported zirconium catalyst is reported.

Reaction of Zr(NMe₂)₄and mesoporous silica nanoparticles

(MSN) provides the catalytic material Zr(NMe₂)_n@MSN.

Exhaustive characterization of Zr(NMe₂)_n@MSN with solid-

state (SS)NMR and infrared spectroscopy, as well as through

reactivity studies, suggests its surface structure is primarily 

SiOZr(NMe₂)₃. The presence of these nitrogen-containing

zirconium sites is supported by ¹⁵N NMR spectroscopy,

including natural abundance ¹⁵N NMR measurements using

dynamic nuclear polarization (DNP) SSNMR. The Zr-

(NMe₂)_n@MSN material reacts with pinacolborane (HBpin) to provide Me₂NBpin and the material ZrH/Bpin@MSN that

is composed of interacting surface-bonded zirconium hydride and surface-bonded borane SiOBpin moieties in an

approximately 1:1 ratio, as well as zirconium sites coordinated by dimethylamine. The ZrH/Bpin@MSN is characterized by

¹H/²H and ¹¹B SSNMR and infrared spectroscopy and through its reactivity with D₂. The zirconium hydride material or the

zirconium amide precursor Zr(NMe₂)_n@MSN catalyzes the selective hydroboration of aldehydes and ketones with HBpin in the

presence of functional groups that are often reduced under hydroboration conditions or are sensitive to metal hydrides, including

oleﬁns, alkynes, nitro groups, halides, and ethers. Remarkably, this catalytic material may be recycled without loss of activity at

least eight times, and air-exposed materials are catalytically active. Thus, these supported zirconium centers are robust catalytic

sites for carbonyl reduction and that surface-supported, catalytically reactive zirconium hydride may be generated from

zirconium-amide or zirconium alkoxide sites.

KEYWORDS: single-site catalysts, carbonyl hydroboration, zirconium hydride, interfacial catalysis, mesoporous silica, solid-state NMR

versions as well as a straightforward means for recycling the

catalytic materials.

INTRODUCTION

■

Surface-supported early-transition-metal hydrides^1,2are highly

reactive toward C−H bond breaking reactions that allow

stoichiometric methane metalation³and catalytic conversions

such as oleﬁn polymerization,⁴hydrogenation,⁵H/D ex-

change,⁶alkane metathesis, and hydrogenolysis of polyethylene

and other alkanes.⁷⁻¹⁰Although the surface provides kinetic

stabilization of metal hydrides against multimetallic decom-

position reactions, solution-phase early-metal and rare-earth

hydrides are implicated in a range of catalytic chemistry

including hydrosilylation,¹¹⁻¹⁷hydrogenation,¹⁸dehydrocou-

pling and dehydrogenative polymerization,¹⁹⁻²¹and hydro-

boration.^22,23Often, the metal hydrides in these reactions are

generated and used in situ or are proposed as intermediates in

catalytic cycles. This in situ generation could also be an

advantageous approach for the application of surface-supported

metal hydrides in catalysis. Moreover, a surface-supported

hydride such as (SiO)₃ZrH could tolerate harsher conditions

(e.g., higher temperature) than soluble analogues in catalytic

addition chemistry, thereby permitting more diﬃcult con-

We recently reported a homogeneous magnesium-catalyzed

cleavage and hydroboration of esters using an in situ generated

magnesium hydridoborate catalyst.²⁴Despite the oxophilicity of

the magnesium center, the catalytic site could be generated by

reaction of pinacolborane (HBpin) and magnesium alkoxide.

Similarly, [{Nacnac}MgH]₂(Nacnac = ((2,6-iPr₂C₆H₃)-

NCMe)₂HC) is a highly active catalyst for hydroboration of

pyridines, ketones, and aldehydes.^22,25A related zwitterionic

magnesium catalyst is suﬃciently reactive to reduce carbon

dioxide to a methanol equivalent.²⁶Hydroboration of aldehydes

and ketones is catalyzed by soluble titanium,²⁷molybdenum,²⁸

as well as a few late metal catalysts.^29,30Recently, divalent

germanium and tin compounds were also shown to catalyze this

reaction, and hydrides were postulated intermediates.³¹Group

4-catalyzed carbonyl hydrosilylations and hydrogenations are

Received: July 31, 2015

Revised: October 29, 2015

7399

DOI: 10.1021/acscatal.5b01671

ACS Catal. 2015, 5, 7399−7414

ACS Catalysis

Research Article

also known,^11,32−35and although Schwartz’s reagent catalyzes

hydroboration of alkynes,³⁶we are not aware of previous

reports of zirconium-catalyzed hydroboration of carbonyls. In

addition, the reaction of Zr(NMe₂)₄and SBA-15 was recently

reported to give an azazirconocyclopropane surface species,³⁷as

does a titanium amide on silica en route to a titanium imido.³⁸

Notably, the azazirconocyclopropane species reacts with

hydrogen to give a zirconium hydride that catalyzes hydro-

genation of oleﬁns.³⁹The catalytic C−H and C−C bond-

breaking and -forming reactions of surface-supported zirconium

hydride are notoriously sensitive to oxygen-containing

impurities, which give irreversible deactivation of the catalytic

sites.

As a possible solution to these deactivation processes, we

envisioned that surface-supported oxophilic metal complexes

with oxygen- or nitrogen-containing ligand precursors could be

activated with reducing reagents such as boranes. Such

transformations could potentially allow access to highly reactive

surface-supported zirconium hydride sites under mild con-

ditions and also provide a means for reactivating deactivated

catalytic sites. On the basis of these ideas and the known

chemistry of (SiO)₃ZrH,^1,3,40,41the hydroboration of

carbonyls appears to be an appropriate choice for testing the

surface-supported Zr−OR bond cleavage steps in a catalytic

cycle. Reductions of ketones and aldehydes are readily achieved

with stoichiometric boron-containing reagents such as BH₃·

THF or NaBH₄or highly reactive metal hydrides such as

LiAlH₄.^42,43However, selectivity for carbonyls versus oleﬁns

and other functional groups including organohalides and nitro

groups are limited with these reagents, these reagents are easily

hydrolyzed by adventitious moisture, and their reactions

produce substantial amounts of salt waste. Thus, alternative

catalytic methods for selective carbonyl reductions, employing

earth-abundant catalysts, are desirable. In this context,

supported single-site hydroboration catalysts based on earth

abundant zirconium would represent a signiﬁcant conceptual

advance in the ﬁeld. In addition, a heterogeneous catalyst could

oﬀer advantages in sustainable synthesis through recyclable

catalytic materials and in ﬂow chemistry.

N₂sorption/desorption, powder XRD, TEM, solid-state H,

1

²H, ¹¹B, ¹³C, and ¹⁵N SSNMR spectroscopy, ¹⁵N SSNMR

spectroscopy enhanced by dynamic nuclear polarization

(DNP), and infrared spectroscopy. Zr(NMe₂)₄,⁴⁵{PhB-

(Ox^Me2)₂C₅H₄}Zr(NMe₂)₂(Ox^Me2= 4,4-dimethyl-2-oxazo-

line),⁴⁶Si(SiMe₃)₄,^47,48and DBpin⁴⁹were synthesized

according to literature procedures. Pinacolborane (used as

received) and 98% ¹⁵N-labeled [H₂¹⁵NMe₂]Cl (dried under

vacuum at 120 °C for 2 h) were purchased from Aldrich.

Solution-phase ¹H, ¹³C{¹H} and ¹¹B NMR spectra were

collected either on a Bruker DRX 400 MHz spectrometer,

Bruker Avance III 600 MHz spectrometer or a Varian MR 400

MHz spectrometer. ¹¹B NMR spectra were referenced to an

external sample of BF₃·Et₂O. Infrared spectra were recorded on

neat MSN samples using a Bruker Vertex 80 spectrometer with

a Harrick Praying Mantis Diﬀuse Reﬂection Accessory in a

reaction chamber with ZnSe windows. These samples were

prepared and maintained under an inert N₂atmosphere.

Elemental analyses were performed using a PerkinElmer 2400

Series II CHN/S in the Iowa State Chemical Instrumentation

Facility.

Inductively coupled plasma-optical emission spectroscopy

(ICP-OES) was performed on 10 samples to measure the

zirconium loading in Zr(NMe₂)_n@MSN and zirconium and

boron loading in ZrH/Bpin@MSN. The samples (2.0−4.0 mg

each) were digested for 24 h in aqueous HF and HCl solution

(0.18% and 5% respectively) and analyzed in a PerkinElmer

Optima 2100 DV ICP-OES instrument.

N₂sorption isotherms were measured in a Micromeritics

Tristar surface area analyzer. Samples were previously purged

for 6 h under a N₂ﬂow at 393 K, and the isotherms were

determined at 77 K. The surface area was calculated using the

Brunauer−Emmett−Teller equation, and the pore size

distribution was obtained from analysis of the adsorption

branch of the isotherm using the Barrett−Joyner−Halenda

method.

Powder X-ray diﬀraction (XRD) patterns were obtained with

a Rigaku Ultima IV diﬀractometer using a Cu target at 40 kV

and 44 mA. Kβ was removed with a monochromator, and the

data were collected from 0.7 to 8 2θ° with a resolution of 0.02

2θ°.

Transmission electron microscopy (TEM) and high-angle

annular dark ﬁeld scanning TEM (HAADF-STEM) images

were acquired in a Tecnai G2 F20 electron microscope

operated at 200 kV. Samples were prepared by dispersion into

benzene, deposition of a single drop in a copper grid coated

with lacey carbon, and evaporation at room temperature.

Energy-dispersive X-ray (EDX) spectra were collected on

representative areas to probe for homogeneity of elemental

composition.

Here we report the synthesis and characterization of a

mesoporous silica nanoparticle (MSN)-supported zirconium

amide complex identiﬁed as Zr(NMe₂)_n@MSN. In the ﬁrst

section of this paper, we describe the details supporting the

assigned surface structures of this material. In the second

section, we describe the reactivity of Zr(NMe₂)_n@MSN with

pinacolborane and the nature of the surface species ZrH/

Bpin@MSN produced from that reaction. Finally, we present

the catalytic activity of Zr(NMe₂)_n@MSN as a recyclable

catalyst for the reduction of carbonyls by catalytic hydro-

boration.

METHODS AND MATERIALS

SSNMR measurements were performed on a 600 MHz

Varian NMR System spectrometer, equipped with a 1.6 mm

magic-angle spinning (MAS) probe. Several one-dimensional

(1D) and two-dimensional (2D) experiments were used,

including 1D ¹H, ²H, ¹³C, and ¹¹B MAS with direct polarization

■

General. All reactions were performed under a dry argon

atmosphere using standard Schlenk techniques or under a

nitrogen atmosphere in a glovebox, unless otherwise indicated.

Dry, oxygen-free solvents were used throughout. Benzene,

toluene, pentane, methylene chloride, and tetrahydrofuran were

degassed by sparging with nitrogen, ﬁltered through activated

alumina columns, and stored under nitrogen. Benzene-d₆was

heated to reﬂux over Na/K alloy and vacuum-transferred. SBA-

15 type MSN was synthesized according to the literature,⁴⁴

calcined at 550 °C, washed with water, and then heated to 550

°C under vacuum. The catalytic materials were characterized by

1

(DPMAS), 1D H → ¹³C cross-polarization under MAS (¹³C

CPMAS), ¹⁵N CPMAS, ¹¹B CPMAS, 2D ¹¹B triple-quantum

1

(3Q)MAS, as well as 2D H−¹¹B heteronuclear correlation

(Hetcor) NMR and indirectly detected ¹⁵N−¹H (id)Hetcor

NMR. The samples were packed in zirconia MAS rotors in a

glovebox under nitrogen atmosphere. NMR experiments were

carried out under N₂atmosphere, as well.

7400

DOI: 10.1021/acscatal.5b01671

ACS Catal. 2015, 5, 7399−7414

ACS Catalysis

Research Article

The ¹⁵N DNP-enhanced CPMAS experiments were

performed on a 400 MHz Bruker DNP SSNMR spectrometer

equipped with a low-temperature (∼100 K) MAS probe. The

sample was prepared by impregnating the MSNs with a 16 mM

solution of TEKPol in 1,1,2,2-tetrachloroethane (predried by

stirring overnight with CaCl₂followed by distillation under N₂),

and then packed into a 3.2 mm sapphire MAS rotor.^50,51

The SSNMR experimental parameters are given in the ﬁgure

captions using the following symbols: ν_Rdenotes the MAS rate,

ν_RF(X) is the magnitude of the RF ﬁeld applied to X nuclei, τ_CP

is the cross-polarization contact time, τ_RDis the recycle delay,

and Δt₁is the time interval of t₁during 2D acquisition. ¹H/¹³C

and ¹⁵N chemical shifts were referenced to TMS and

nitromethane as 0 ppm, respectively.

white solid (0.253 g). IR (KBr, cm⁻¹): 2852 (m), 2777 (m),

1457 (m), 1084 (s, ν_Si−O), 950 (m). Elemental analysis: Found:

C, 5.91; H, 1.08; N, 3.44; Zr, 8.3 wt % (0.91 mmol).

Bpin@MSN. Pinacolborane (0.075 g, 0.594 mmol) dissolved

in benzene was added to a suspension of calcined MSN (0.20 g,

0.34 mmol of SiOH) in benzene (5 mL). Vigorous bubbling

was observed immediately. No more bubbling was observed

after 2 h of stirring, the mixture was centrifuged, and the

solvent was decanted. The unreacted HBpin was removed from

the solid material by washing with benzene (3 × 5 mL) and

then pentane (2 × 5 mL). The solid material was dried under

reduced pressure yielding a white solid (0.226 g). IR (KBr,

cm⁻¹): 2985 (m), 2938 (w), 1480 (m), 1456 (w), 1375 (m),

1223 (m), 1156 (m), 1086 (s, ν_Si−O), 950 (m). Elemental

analysis: Found: C, 9.99; H, 0.98; N, 0.03; B, 14.3 wt % (1.33

mmol).

ZrH/Bpin@MSN. Pinacolborane (0.691 g, 5.40 mmol)

dissolved in benzene was added to Zr(NMe₂)_n@MSN (0.200

g, 0.182 mmol of Zr, 0.540 mmol of NMe₂) suspended in

benzene (5 mL). Slow evolution of a small amount of bubbles

was observed, and this bubbling was signiﬁcantly reduced

compared to the Bpin@MSN sample. The mixture was stirred

at 60 °C for 2 h, allowed to settle in a centrifuge, and then the

solvent was decanted. The unreacted HBpin and Me₂NBpin

were removed from the solid material by washing with benzene

(3 × 5 mL) and pentane (2 × 5 mL). The solid material was

dried under reduced pressure yielding a white solid (0.207 g).

IR (KBr, cm⁻¹): 2979 (m), 2934 (w), 2805 (w), 1592 (w, Zr−

H), 1479 (m), 1376 (w), 1095 (s, ν_Si−O), 950 (m). Elemental

analysis: Found: C, 8.09; H, 1.00; N, 0.51; Zr, 8.1 wt % (0.89

mmol); B, 9.3 wt % (0.86 mmol). Companion in situ

micromolar scale reactions were performed in a J. Young-

style Teﬂon-sealable NMR tube with 0.013 g Zr(NMe₂)_n@

MSN, 0.041 g HBpin, and benzene-d₆as solvent with a 7.48

mM Si(SiMe₃)₄standard. From the integrated values of the

Me₂NBpin and Si(SiMe₃)₄resonances, 0.102 mmol of

Me₂NBpin was formed.

Zr(NMe₂)₄. Labeled H¹⁵NMe₂was only available as

Me₂¹⁵NH·HCl, so the synthesis of Zr(NMe₂)₄from Me₂NH·

HCl was developed, ﬁrst with unlabeled starting material and

then on small scale with the isotopically enriched material.

Dried Me₂NH·HCl (0.500 g, 6.133 mmol) was suspended in

tetrahydrofuran (50 mL) and cooled to −78 °C, and nBuLi

(4.9 mL, 12.3 mmol) was added. The mixture was stirred at this

temperature for 2 h and then warmed to room temperature and

stirred for 12 h. The volatile components were evaporated

under reduced pressure to give a solid residue, which was

washed with pentane (3×) and dried under vacuum to yield a

white solid mixture of LiNMe₂, LiCl and a substoichiometric

1

amount of coordinated tetrahydrofuran. H NMR (THF-d₈,

600 MHz): δ 3.63 (br, 1 H, THF), 2.70 (s, 6 H, NMe₂), 1.77

(br, 1 H, THF). ¹³C{¹H} NMR (THF-d₈, 151 MHz): δ 68.3

(THF), 49.1 (NMe₂), 26.5 (THF). ¹⁵N{¹H} NMR (THF-d₈,

59.2 MHz): δ −375.9.

The mixture of LiNMe₂(0.370 g, 3.332 mmol of LiNMe₂)

and LiCl was suspended in toluene. ZrCl₄(0.110 g, 0.472

mmol) was added at room temperature, and the reaction

mixture was stirred for 12 h. The solution was ﬁltered, and the

solvent was evaporated under reduced pressure to yield

1

Zr(NMe₂)₆Li₂THF₂(0.231 g, 0.449 mmol, 96%). H NMR

General Procedure for the Catalytic Hydroboration of

Carbonyls (Aldehydes and Ketones) Using Zr(NMe₂)_n@

MSN. A mixture of the carbonyl substrate (1 mmol) and

HBpin (1.3 mmol) was added to a Zr(NMe₂)_n@MSN (0.05

mmol Zr) that was suspended in benzene (10 mL). The

reaction mixture was stirred for 10 h at room temperature for

aldehydes or at 60 °C for ketones. The reaction mixture was

ﬁltered, the ﬁltrate was quenched with 1 M aqueous NaOH

solution, and the alcohol product was extracted with diethyl

ether. The pure alcohol was obtained after drying the Et₂O

solution over Na₂SO₄for 2 h and evaporating the solvent under

reduced pressure.

(benzene-d₆, 600 MHz): δ 3.35 (br, 8 H, THF), 3.19 (s, 36 H,

NMe₂), 1.18 (br, 8 H, THF). ¹³C{¹H} NMR (benzene-d₆, 151

MHz): δ 68.8 (THF), 46.9 (NMe₂), 25.7 (THF). ¹⁵N{¹H}

NMR (benzene-d₆, 59.2 MHz): δ −295.1.

Zr(NMe₂)₆Li₂THF₂(0.200 g, 0.390 mmol) was dissolved in

benzene. ZrCl₄(0.045 g, 0.195 mmol) was added at room

temperature, and the reaction mixture was stirred for 10 min.

The solution was ﬁltered, and the solvent was evaporated under

vacuum to yield Zr(NMe₂)₄(0.180 g, 0.673 mmol, 86%). The

¹H and ¹³C{¹H} NMR spectra matched the reported literature

values.⁵²¹H NMR (benzene-d₆, 600 MHz): δ 2.97 (s, NMe₂).

Zr(¹⁵NMe₂)₄. The above procedure, employing Me₂¹⁵NH·

HCl (0.123 g, 1.12 mmol) and ZrCl₄(0.036 g, 0.155 mmol),

aﬀorded Zr(¹⁵NMe₂)₆Li₂THF₂(0.074 g, 0.142 mmol, 93%).

Reaction of this material with ZrCl₄(0.033 g, 0.141 mmol)

RESULTS AND DISCUSSION

■

Synthesis and Characterization of Zr(NMe₂)_n@MSN.

Tetrakis(dimethylamido)zirconium was grafted on high surface

area mesoporous silica to give Zr(NMe₂)_n@MSN. The silica

support, in the form of SBA-15 type MSN characterized by a

hexagonal array (p6mm) of 9.7 nm diameter pores and a surface

area of 385 m²/g, was produced by hydrolysis−condensation of

tetramethylorthosilicate using the Pluronic P104 template,

calcined at 550 °C, washed with water, then heated at 550 °C

under vacuum, and subsequently stored in a glovebox away

from ambient air and moisture.⁴⁴The SiOH group surface

concentration of 1.7 mmol/g was determined by measuring the

concentration of toluene produced in a titration with

1

provided Zr(¹⁵NMe₂)₄(0.061 g, 0.225 mmol, 79%). H NMR

(benzene-d₆, 600 MHz): δ 2.98 (s, NMe₂). ¹⁵N{¹H} NMR

(benzene-d₆, 59.2 MHz): δ −306.2.

Zr(NMe₂)_n@MSN. A benzene solution of Zr(NMe₂)₄(0.095

g, 0.355 mmol, 5 mL) was added to MSN (0.20 g, 0.34 mmol

of SiOH groups) suspended in benzene (15 mL). The

suspension was stirred for 20 h at ambient temperature, the

mixture was centrifuged, and the solvent was decanted. The

unreacted Zr(NMe₂)₄was removed from the solid material by

washing with benzene (3 × 5 mL) and then pentane (2 × 5

mL). The solid material was dried under vacuum yielding a

7401

DOI: 10.1021/acscatal.5b01671

ACS Catal. 2015, 5, 7399−7414

ACS Catalysis

Research Article

53

Mg(CH₂Ph)₂(O₂C₄H₈)₂and by spin counting of Q³-sites

with ²⁹Si DPMAS NMR spectroscopy (1.6 mmol/g). Thus-

prepared MSN and Zr(NMe₂)₄react in benzene at room

temperature for 20 h producing Zr(NMe₂)_n@MSN, a grafted

material that we will contend is primarily monopodal

tris(dimethylamido)zirconium, with smaller amounts of bipodal

species bis(dimethylamido)zirconium and bis(dimethylamido)

(dimethylamino)zirconium (eq 1).

There was no evidence for large zirconium clusters formed in

the grafting experiments in the images produced by TEM and

HAADF-STEM (Figure 1B). Likewise, the EDX analysis

(Figure 1C) suggested that zirconium is well dispersed over

the silica. In addition, a powder XRD measurement of

Zr(NMe₂)_n@MSN showed diﬀraction peaks assigned to the

periodic channels of the mesoporous silica support (see

Supporting Information (SI) Figure S1).

Next, the amount of Zr(NMe₂)₄grafted onto MSN was

approximated by quantifying the soluble zirconium amide

before and after its reaction with the silica. A benzene solution

of excess Zr(NMe₂)₄(1.78 mmol) stirred with 1 g of MSN

resulted in the consumption of 1 mmol of Zr(NMe₂)₄,

indicating that the loading is ca. 1 mmol Zr per gram of

MSN. In this reaction, approximately 1.2 mmol of HNMe₂was

produced per gram of silica. These amounts were determined

1

by integration of the reactant and product resonances in H

Upon scale-up, the Zr(NMe₂)_n@MSN is puriﬁed from excess

Zr(NMe₂)₄by pentane and benzene washes. The structural

morphology of the material was characterized by powder XRD

and TEM. The zirconium sites were identiﬁed and charac-

terized by the mass-balance implied by stoichiometry from the

synthesis, quantitative metals analysis using ICP-OES,

combustion analysis, infrared spectroscopy, and SSNMR

spectroscopy, as well as the stoichiometry and observed

products of its reactions with HBpin and D₂.

NMR spectra of the reaction mixtures, which contained a

known concentration of Si(SiMe₃)₄as an internal standard.

This loading was further supported by ICP-OES analysis that

indicated the presence of 8.4 0.1 wt % Zr in Zr(NMe₂)_n@

MSN (0.91 0.1 mmol Zr/g; Table 1). The ICP-OES analysis

involved 10 measurements performed over several days on

samples of Zr(NMe₂)_n@MSN and established the stability of

the zirconium-supported material and the reproducibility of the

method for the comparison of the series of materials derived

from Zr(NMe₂)_n@MSN. Heating of Zr(NMe₂)_n@MSN at 60

°C in benzene did not aﬀect the material’s weight %.

A TEM image (Figure 1A) of Zr(NMe₂)_n@MSN showed

that the ordered mesoporous nature of the SBA-15-type

material is maintained after its treatment with Zr(NMe₂)₄.

Figure 1. (A) TEM of Zr(NMe₂)_n@MSN, (B) HAADF-STEM of Zr(NMe₂)_n@MSN, (C) EDX analysis of the region enclosed in the red square in

the HAADF-STEM image of Zr(NMe₂)_n@MSN, (D) TEM of Zr(NMe₂)_n@MSN + HBpin, (E) HAADF-STEM of Zr(NMe₂)_n@MSN + HBpin,

and (F) EDX analysis of the regions enclosed in the red squares in the HAADF-STEM image showing consistent Zr:Si ratios throughout the

material.

7402

DOI: 10.1021/acscatal.5b01671

ACS Catal. 2015, 5, 7399−7414

ACS Catalysis

Research Article

Table 1. Zirconium Loading of Zr(NMe₂)_n@MSN Materials Obtained by ICP-OES Analysis

a

b

material preparation

conditions

weight % Zr

mmol Zr/g

Zr(NMe₂)₄+ MSN → Zr(NMe₂)_n@MSN

Zr(NMe₂)_n@MSN heated at 60 °C

20 h, r.t.

2 h, 60 °C

8.4 0.1

8.2 0.1

0.91

0.92

0.89

Zr(NMe₂)_n@MSN + 10 HBpin → ZrH/Bpin@MSN

Zr(NMe₂)_n@MSN + 10 HBpin +10 PhMeCO

a

b

Benzene solvent. Determined by ICP-OES analysis.

Figure 2. Diﬀuse reﬂectance infrared spectra of (A) MSN, (B) Zr(NMe₂)₄, (C) Zr(NMe₂)_n@MSN, and (D) MSN + HNMe₂.

The Zr:NMe₂ratio in Zr(NMe₂)_n@MSN, as well as the

carbon (5.91%, 4.9 mmol/g), hydrogen (1.08%, 10.8 mmol/g),

and nitrogen (3.44%, 2.5 mmol/g) loadings were measured

using combustion analysis. From these data, the Zr:C:N ratio is

1:5.4:2.7 corresponding to a Zr:NMe₂ratio of 1:2.7. These

results, corroborated by the measurements of the stoichiometry

of the grafting reactions, imply that the material contains

primarily zirconium sites with three nitrogen-containing ligands

and a smaller amount (up to 30%) of sites with two nitrogen-

containing ligands.

The presence of NMe₂groups in Zr(NMe₂)_n@MSN was

identiﬁed through infrared spectroscopy. An IR spectrum of

MSN (calcined, washed with water, and then dried at 550 °C

under vacuum) contained an absorption at 3747 cm⁻¹assigned

to isolated SiOH groups (Figure 2A).⁵⁴Also, for comparison,

the infrared spectrum of Zr(NMe₂)₄contained bands at 2942,

2869, 2773, and 1457 cm⁻¹(Figure 2B; see Figures S2 and S3

for full IR spectra).

After the reaction between MSN and Zr(NMe₂)₄, several

new signals associated with organic groups were observed at

2852, 2777, and 1457 cm⁻¹. These signals are similar to those

observed for Zr(NMe₂)₄, as can be seen through comparison of

Figure 2B,C. After the grafting reaction, the IR band associated

with isolated surface SiOH moieties was diminished, and it is

likely that some unreacted SiOH groups are still present on the

surface. These groups, as well as likely NH containing species,

are diﬃcult to detect as a result of their low concentration and

possible broadening due to hydrogen bonding.

under solid−gas conditions (Figure 2D) in three independent

experiments, followed by evacuation. In all cases, a small signal

at 1457 cm⁻¹and even weaker intensity signals typically

attributed to ν_CHor ν_NH(at 3334 cm⁻¹) were observed in the

infrared spectra. The peak at 3747 cm⁻¹assigned to isolated

SiOH groups was observed after the HNMe₂treatments. In

addition, a weak, yet sharp signal was observed in the ¹³C

CPMAS spectrum (not shown), with the chemical shift very

close to neat dimethylamine (∼35 ppm) and the intensity

corresponding to less than 0.1 mmol/g. From these experi-

ments, we conclude that only a small amount of HNMe₂

associates with the MSN material in physisorbed form.

Moreover, these sites may be blocked once zirconium amide

is grafted to the silica surface.

The ¹³C CPMAS SSNMR spectrum of Zr(NMe₂)_n@MSN

(Figure 3, top) showed two strongly overlapping signals with

the chemical shifts of ∼36 and ∼39 ppm, which are similar to

that of Zr(NMe₂)₄dissolved in benzene-d₆(42 ppm). No other

resonances were detected, even after 76 000 acquisitions. The

completeness of the CPMAS spectrum was conﬁrmed by the

¹³C DPMAS experiment (Figure 3, bottom), which yielded the

same line shape. These ¹³C spectra contrast with those reported

earlier by El Eter et al., in which SiOZr(η²-CH₂NMe)

(NMe₂)(NHMe₂), formed from the reaction of SBA-15₇₀₀(i.e.,

mesoporous silica pretreated at 700 °C) and Zr(NMe₂)₄in

pentane for 1 h produced three signals at 36, 47, and 85 ppm.³⁷

DFT calculations showed that the chemical shifts of all

methyl carbons in dimethylamido zirconium model surface

moieties, including SiOZr(NMe₂)₃and (SiO)₂Zr-

(NMe₂)₂, are expected between 36 and 43 ppm (see Figure

S4 in SI), strongly supporting the hypothesis that both

The HNMe₂reaction byproduct might be expected to

interact with the silica surface. To test for this possibility, MSN

and HNMe₂were allowed to react in benzene, pentane, or

7403

DOI: 10.1021/acscatal.5b01671

ACS Catal. 2015, 5, 7399−7414

ACS Catalysis

Research Article

(NMe₂)₂(up to 30%) is likely to be present. The former

species may be either the monopodal SiOZr(NMe₂)₃or a

dipodal diamido amine adduct (SiO)₂Zr(NMe₂)₂(NHMe₂).

The monopodal stoichiometry would imply that ∼0.7−0.8

mmol/g of SiOH remained intact, and indeed, the

experiments with pinacolborane described below suggest that

accessible SiOH groups remain on the silica surface. In

contrast, the diﬀuse reﬂectance FTIR spectrum does not

contain signals in the expected region for isolated SiOH. That

is, neither the IR peak at 3747 cm⁻¹, associated with isolated

silanols, nor a broad signal for hydrogen-bonded silanols are

observed. However, a weak signal at 3292 cm⁻¹may be

assigned to an NH stretching band of a possible zirconium-

coordinated dimethylamine, and this signal is slightly shifted

from the signal of physisorbed HNMe₂on MSN. On the basis

of the residual nitrogen loading after treatment of Zr(NMe₂)_n@

MSN with HBpin (see below), the amount of zirconium-

coordinated dimethylamine is estimated to be less than 10%.

Because SiOZr(NMe₂)₃, (SiO)₂Zr(NMe₂)₂(NHMe₂),

or (SiO)₂Zr(NMe₂)₂are not conclusively distinguished as

surface structures by ¹³C SSNMR, IR, and elemental analysis,

we turned to ¹⁵N SSNMR measurements to further characterize

the nitrogen-containing ligands bonded to zirconium. At

natural ¹⁵N abundance, NMR signals could not be detected

either in 1D ¹⁵N spectra or in 2D ¹H−¹⁵N correlation

experiments. We thus resorted to the newly developed DNP

method, which enhances the sensitivity of SSNMR of surface

species by ∼2 orders magnitude via excitation of the

exogenously introduced biradicals (here TEKPol dissolved in

1,1,2,2-tetrachloroethane) at their ESR resonance frequency,

followed by transfer of magnetization to the nuclear spins.^50,51

A high-quality DNP-enhanced ¹⁵N CPMAS spectrum of

Zr(NMe₂)_n@MSN was indeed acquired under natural

Figure 3. ¹³C CPMAS (top) and DPMAS (bottom) spectra of

Zr(NMe₂)_n@MSN obtained under 25 kHz MAS with ¹H TPPM

heteronuclear decoupling at ν_RF(¹H) = 100 kHz. The CPMAS

spectrum was measured using ν_RF(¹H) = 75 kHz and ν_RF(¹³C) = 100

kHz during CP, τ_CP= 4.5 ms, τ_RD= 1.5 s, and 76 000 scans. The

DPMAS spectrum resulted from 360 scans with τ_RD= 60 s.

resonances observed in Zr(NMe₂)_n@MSN represent the Zr-

bound NMe₂functionalities. As detailed in the SI, the DFT

calculations were based on the gauge-including projector-

augmented wave (GIPAW) method.⁵⁵The observed non-

equivalence of methyl groups can be attributed to diﬀerences in

local geometries and mobilities (see eq 1).

Spin counting, using a ¹³C DPMAS NMR experiment,

quantiﬁed the NMe₂loading in Zr(NMe₂)_n@MSN at 2.7

( 0.5) mmol/g. Considering the fact that the Zr loading

measured with ICP-OES was 0.91 mmol/g, the Zr:NMe₂ratio

is estimated at close to 1:3. This value and the value obtained

from combustion analysis (1:2.7) described above are in

suﬃcient agreement to suggest the surface species in Zr-

(NMe₂)_n@MSN primarily comprises three nitrogen-containing

ligands, although some quantity of bipodal (SiO)₂Zr-

Figure 4. (A) Top spectrum: DNP-enhanced ¹⁵N CPMAS spectrum of Zr(NMe₂)_n@MSN under natural ¹⁵N abundance. The spectrum was

1

obtained at ∼100 K using ν_R= 10 kHz; ν_RF(¹H) = 100 kHz, 107 kHz, and 100 kHz during hard pulse, cross-polarization, and SPINAL H

decoupling; ν_RF(¹⁵N) = ∼ 87 kHz during cross-polarization; τ_CP= 4 ms; 2048 scans; and τ_RD= 4 s. Lower spectrum: skyline ¹⁵N projection of the 2D

¹⁵N−¹H idHetcor spectrum in ﬁgure (B). (B) 2D ¹⁵N−¹H idHetcor spectrum of ¹⁵N-enriched Zr(¹⁵NMe₂)_n@MSN. The spectrum was obtained at

14.1 T, using ν_R= 34 kHz; ν_RF(¹H) = 100 kHz during 90° pulse and 91 kHz during CP, and 10 kHz during SPINAL-64 ¹H decoupling; ν_RF(¹⁵N) =

62 kHz during 90° pulse and 57 kHz during CP, and 10 kHz during SPINAL-64 ¹⁵N decoupling; 128 rows with Δt₁= 30 μs; 64 scans per row, and

STATES-TPPI acquisition with τ_RD= 2 s.

7404

DOI: 10.1021/acscatal.5b01671

ACS Catal. 2015, 5, 7399−7414

ACS Catalysis

Research Article

abundance in just over 2 h (Figure 4A, top spectrum). The

spectrum features a single, fairly broad signal at −355 ppm,

which we assign to Zr-bound NMe₂groups. To further

Zr(¹⁵NMe₂)_n@MSN and H¹⁵NMe₂(¹⁵N NMR in benzene-d₆:

−366 ppm). A 2D ¹⁵N−¹H correlation spectrum of Zr-

(¹⁵NMe₂)_n@MSN was acquired under fast (34 kHz) MAS

using the indirect detection of ¹⁵N nuclei for sensitivity

enhancement (¹⁵N−¹H idHetcor, see Figure 4B).⁵⁷In

agreement with the DNP-based experiment, the idHetcor

investigate their nature, a series of DNP-enhanced CPMAS ¹⁵

N

NMR spectra were measured as a function of τ_CPcontact time.

The buildup of H−¹⁵N cross-polarization (Figure S5 in SI),

1

which is governed by the heteronuclear dipolar coupling, and

measurement showed a dominant correlation between the H

56

thus can be used to evaluate the H−¹⁵N distance, indicates

1

NMR signal at ∼2.4 ppm and NMe₂groups resonating at

around −355 ppm. A minor peak at −370 ppm most likely

represents small concentration of free HNMe₂left within the

pores.

that the ¹⁵N nuclei are polarized primarily by H nuclei at a

1

distance of about 2 Å, which is consistent with protons in

NMe₂groups being the source. Interestingly, the local maxima

in the curve (which were reproducibly measured several times,

and evaluated using eq 20c in ref 49), suggest that a small

In addition, note that Zr(NMe₂)_n@MSN features close to 1

mmol of functional groups per 385 m²of surface, which

corresponds to the coverage of ∼70%. Thus, the grafting

reaction provides the maximum zirconium amide loading. We

may further speculate that the distinction between 

SiOZr(NMe₂)₃, (SiO)₂Zr(NMe₂)₂(NHMe₂), and (

SiO)₂Zr(NMe₂)₂may relate to steric eﬀects controlled by

this surface coverage. The distinction between the surface

species obtained in our grafting experiments versus the

cyclometalated SiOZr(η²-NMeCH₂)(NMe₂)(NHMe₂) may

1

fraction of the H−¹⁵N pairs reside at a distance of ∼1.0 Å, as

would be expected in zirconium amine species. This ﬁnding

implies that the resonance centered at −355 ppm can be

assigned to dimethylamide groups and a small amount of

dimethylamine coordinated to surface-bonded zirconium sites.

We decided to verify that the surface Zr species observed by

DNP did not result from the reaction with the solvent or the

nitroxide radicals. To this end, we synthesized labeled

Zr(¹⁵NMe₂)₄from isotopically enriched Me₂¹⁵NH·HCl through

the sequence shown in Scheme 1. The intermediate species

also relate to these intersite steric eﬀects. In particular, silica₇₀₀

,

which is dehydroxylated at 700 °C under vacuum to give

isolated silanols, reacts to provide only monopodal surface

structures. In the 550 °C calcined MSN used in our study, a

bipodal zirconium-surface interaction relieves the intersite steric

pressure rather than β-abstraction that would give the

zirconacyclopropane structure.

Scheme 1. Synthesis of Labeled Zr(¹⁵NMe₂)₄

The ¹⁵N NMR experiments also rule out the presence of 

SiNMe₂surface groups, which could form by addition of Zr−

NMe₂across a strained Si−O−Si surface site. This conclusion

is based on the featureless region of the ¹⁵N NMR experiments

from −300 to −350 ppm. The ¹⁵N NMR signal for SiNMe₂

is expected to be ca. −330 ppm based on the ¹⁵N NMR

chemical shift of the model compound (EtO)₃SiNMe₂(¹⁵N

NMR, benzene-d₆: −326 ppm). A summary of all ﬁndings that

support the composition of Zr(NMe₂)_n@MSN as a primarily

SiOZr(NMe₂)₃, with smaller amounts of (SiO)₂Zr-

(NMe₂)₂and (SiO)₂Zr(NMe₂)₂(NHMe₂) is given in Table

2.

Zr(NMe₂)₆Li₂THF₂was previously reported from the reaction

of Zr(NMe₂)₄and 2 equiv of LiNMe₂.⁵²Here it is synthesized

directly from ZrCl₄and LiNMe₂, and we report its ¹⁵N NMR

chemical shift at −295 ppm.

The reaction of Zr(¹⁵NMe₂)₆Li₂THF₂and 0.5 equiv of ZrCl₄

gives pure Zr(¹⁵NMe₂)₄as its ¹⁵N labeled isotopomer (¹⁵N

NMR in benzene-d₆: −306 ppm). The grafting reaction was

repeated with the labeled Zr(¹⁵NMe₂)₄sample to produce

Synthesis and Characterization of ZrH/Bpin@MSN.

The reaction of Zr(NMe₂)_n@MSN and HBpin, which produces

Table 2. Characterization Experiments and Conclusions Regarding the Nature of Zr(NMe₂)_n@MSN

experiment

observation

interpretation

no Zr or ZrO₂clusters formed

electron microscopy/

EDX

well-dispersed Zr

reaction stoichiometry

1.2 mmol HNMe₂detected/g MSN

1.0 mmol Zr(NMe₂)₄consumed/g MSN

0.91 mmol Zr/g

ICP-OES

zirconium loading established

C, N combustion analysis 4.9 mmol C/g, 2.5 mmol N/g

Zr:NMe₂∼ 1:2.7 suggests a 3:7 mixture of Zr(NMe₂)₂and

Zr(NMe₂)₃groups

¹³C DPMAS/spin

counting

2.7 mmol NMe₂/g

Zr:NMe₂∼ 1:3 suggests primarily three NMe₂-containing ligands/

Zr

IR

new ν_CHbands at ∼2900 cm⁻¹

NMe₂groups present on surface

ν_OHband at 3747 cm⁻¹not detected

isolated silanols have reacted with Zr(NMe₂)₄

¹⁵N SSNMR:

MSN + HNMe₂, ¹³C

NMR:

strong signal at −355 ppm polarized by Me groups, weakly by H nitrogen is primarily present as dimethylamide groups

weak peak at 36 ppm

only a small amount of HNMe₂physisorbs to calcined MSN

IR analysis:

ν_OHband at 3747 cm⁻¹and weak intensity NH and CH bands are

detected

reaction with HBpin

(below)

2.5 mmol Me₂NBpin formed/g Zr(NMe₂)_n@MSN

ca. 2.7 reactive NMe₂groups per Zr center

7405

DOI: 10.1021/acscatal.5b01671

ACS Catal. 2015, 5, 7399−7414

ACS Catalysis

Research Article

the zirconium hydride surface species discussed below, also

further characterizes the zirconium amide sites by quantiﬁcation

of reactive NMe₂groups (eq 2). A micromolar-scale reaction of

species in Bpin@MSN to be around 21 ppm, which is close to

one measured in the solution NMR spectrum of PhCH₂OBpin

(23 ppm), suggesting that the SiOBpin groups are indeed a

product of the reaction of eq 3. This conclusion is supported by

1

two additional ﬁndings. First, the H DPMAS spectrum of

Bpin@MSN (Figure S6B in SI) features a dominant resonance

1

at 1 ppm, consistent with one observed in the solution H

NMR spectrum of PhCH₂O-Bpin for the pinacol moiety (1.04

ppm). Second, as in the case of ¹⁵N CPMAS, by measuring the

buildup of ¹H →¹¹B CP signal as a function of τ_CP, we

1

estimated the H −¹¹B internuclear distance in Bpin@MSN at

∼3.4 Å (Figure S7A in SI), in good agreement with the average

distance between the ¹¹B and methyl protons in Bpin.⁵⁶

Zr(NMe₂)_n@MSN and excess HBpin aﬀords 2.5 mmol of

Me₂NBpin per gram of material, which was quantiﬁed by

integration of product signals with respect to a known

concentration of Si(SiMe₃)₄as an internal standard. In the

¹H NMR spectrum of the soluble portion of the reaction

mixture, a resonance at 2.62 ppm was assigned to the NMe₂

moiety of Me₂NBpin. In the corresponding solution-phase ¹¹B

NMR spectrum, a singlet at 24.2 ppm was assigned to

Me₂NBpin,⁵⁸and a doublet at 28.5 ppm (¹J_BH= 174 Hz)

represented unreacted HBpin. The reaction of ¹⁵N-labeled

Zr(¹⁵NMe₂)_n@MSN yields an isotopically enriched sample of

Me₂¹⁵NBpin. The ¹⁵N NMR chemical shift of this material

appears at −350 ppm, and this value will be used to assign

surface species (see below). The amount of Me₂NBpin

quantiﬁed by integration suggests that approximately 2.8

NMe₂groups are accessible per zirconium center, again

indicating that the predominant surface species in Zr-

(NMe₂)_n@MSN contains three NMe₂groups per zirconium.

Finally, this experiment rules out the presence of any

cyclometalated amide in this sample because the interaction

of two equivalents of HBpin and the SiOZr(NMeCH₂)-

(NMe₂)(NHMe₂) surface moiety would give pinB−NMeCH₂−

Bpin, and that species was not detected in the solution-phase

¹¹B NMR spectrum.

We now return to ZrH/Bpin@MSN produced by the

reaction given by eq 2. First, ICP-OES measurements show

similar zirconium loading (0.89 mmol/g) and boron loading

(0.86 mmol/g), and that the loading of Zr from Zr(NMe₂)_n@

MSN is unaﬀected by the treatment with HBpin. The ¹¹B spin

counting experiment yielded 0.9 ( 0.1) mmol/g of boron in

ZrH/Bpin@MSN, in excellent agreement with the ICP-OES

analysis. In the ¹¹B DPMAS spectrum of this sample, a broad

signal appeared whose NMR shift (δ ≈ 18 ppm) is similar but

not identical with that of Bpin@MSN discussed above

(compare solid and dashed lines in Figure 5A). The 2D

MQMAS experiment on this sample (Figure 5B), which

removes the anisotropic broadening due to the second-order

quadrupolar interaction and allows for determination of the

pure chemical shift (δ_CS),^59,60shows that the ¹¹B chemical shift

for this boron site is the same as in Bpin@MSN (δ_CS≈ 21

ppm). The so-called isotropic (vertical) dimension of this

spectrum revealed a small shoulder representing an additional

resonance with δ_CS≈ 23 ppm (vide infra), which most likely

represents residual Me₂NBpin trapped within the MSN pores.

1

We also measured the H DPMAS and 2D H−¹¹B Hetcor

1

spectra of ZrH/Bpin@MSN (Figure 5C,D). Note that the H

projection of the Hetcor spectrum is very similar to the H

DPMAS spectrum. The dominant H peak at ∼1 ppm is easily

1

The ZrH/Bpin@MSN sample was synthesized following the

reaction given by eq 2. To facilitate the ensuing discussion of

this product, whose identiﬁcation proved to be very

challenging, we ﬁrst consider an independent reaction 3 of

HBpin on calcined MSN. The ICP measurement of the

resulting material, referred to as Bpin@MSN, indicates a boron

loading of 1.33 mmol/g.

assigned to the protons of the methyl groups of Bpin, whereas

one at ∼2.4 ppm corresponds to ¹H of a small amount of Me₂N

moiety. These results imply that SiOBpin is the dominant

boron-containing structure found in ZrH/Bpin@MSN. How-

ever, in addition to small diﬀerence in the observed shifts

(Figure 5A), the chemical environments of this moiety in

Bpin@MSN and in ZrH/Bpin@MSN samples are not identical.

1

Indeed, the H → ¹¹B CP dynamics indicated that the nearest

¹H−¹¹B internuclear distance is considerably shorter (∼2.1 Å,

Figure S7B in SI) in the zirconium-containing sample. Note

that these experiments do not establish the identity of the

polarizing spin.

The ¹¹B NMR signals for Me₂NBpin and ZrH/Bpin@MSN

are unresolved under MAS alone and thus cannot be discerned

in the Hetcor spectrum. Most likely, the correlation between

the ¹¹B NMR signal and ¹H NMR signal at 2.4 ppm is assigned

to the intermolecular interaction between the ¹¹B of abundant

A

¹¹B DPMAS experiment on Bpin@MSN revealed a single

1

Si−O−Bpin and H of Me₂NBpin, as well as intramolecular

resonance, which, based on the observed NMR shift (δ ≈ 19

ppm, see Figure 5A, dashed line), must be attributed to

trigonally coordinated boron species. We note that the MAS

NMR spectra of half-integer quadrupolar nuclei, such as ¹¹B,

are broadened by the quadrupolar interaction and that the

NMR shifts (δ) observed in such spectra consist of

contributions from the dominant chemical shifts (δ_CS) and

the so-called quadrupole induced shifts. On the basis of the

discussion below, we estimate the δ_CSvalue for the boron

interactions within Me₂NBpin or ZrH/Bpin@MSN species.

Although we propose the surface organometallic species to

1

be a zirconium hydride (eq 3), H NMR resonances at >10

ppm that were previously assigned as (SiO)₃ZrH and (

41

1

SiO)₂ZrH₂were notably absent from the H NMR spectrum

of the product from SiOZr(NMe₂)₃and HBpin. Despite the

absence of those downﬁeld ¹H resonances, a zirconium hydride

species, albeit with a modiﬁed coordination sphere from (

7406

DOI: 10.1021/acscatal.5b01671

ACS Catal. 2015, 5, 7399−7414

ACS Catalysis

Research Article

Figure 5. (A) ¹¹B DPMAS spectra of Bpin@MSN (dashed line) and ZrH/Bpin@MSN (solid line), (B) 2D ¹¹B 3QMAS spectrum of ZrH/Bpin@

MSN, (C) ¹H DPMAS spectrum of ZrH/Bpin@MSN, and (D) 2D ¹H−¹¹B Hetcor spectrum of ZrH/Bpin@MSN. The spectra were obtained using

ν_R= 25 kHz, with (A) ν_RF(¹¹B) = 125 kHz during hard pulse (corresponding to ∼10° ﬂip angle), ν_RF(¹H) = 100 kHz during TPPM ¹H decoupling,

8000 scans, and τ_RD= 1 s; (B) ν_RF(¹¹B) = 125 kHz and 15 kHz during hard and soft (z-ﬁlter) pulses, respectively, ν_RF(¹H) = 100 kHz during TPPM

¹H decoupling, 64 rows with Δt₁= 10 μs, 72 scans per row, and τ_RD= 1.5 s; (C) ν_RF(¹H) = 100 kHz during hard pulse, 4 scans, and τ_RD= 1 s; and

(D) ν_RF(¹H) = 125 kHz, 75 kHz, and 100 kHz during 90° pulse, CP, and TPPM ¹H decoupling, ν_RF(¹¹B) = 50 kHz during CP, τ_CP= 2 ms, 64 rows

with Δt₁= 20 μs, 96 scans per row, and τ_RD= 1 s.

SiO)₃ZrH and (SiO)₂ZrH₂, is a proposed product of the

reaction of Zr(NMe₂)_n@MSN and HBpin. In fact, the room-

temperature, solution−solid interfacial reaction conditions of

the hydroboration are mild with respect to the gas−solid 150

°C reaction of SiOZr(CH₂CMe₃)₃that gives (SiO)₃ZrH

and (SiO)₂ZrH₂. That is, HBpin as a hydride source may

provide access to new zirconium-hydride surface structures.

Unfortunately, we are unaware of any reliable chemical shift

information on structures such as SiOZrH₃; however, on the

basis of the chemical shift trend for (SiO)₃ZrH and (

2

disappeared entirely (Figure 6). Interestingly, the H NMR

SiO)₂ZrH₂, the ¹H NMR signal for SiOZrH₃might be

spectrum from the reaction of ZrH/Bpin@MSN and D₂gives

only a signal at ∼2.5 ppm (Figure S8), suggesting that

deuterium is incorporated in methyl groups. Unfortunately, a

expected to be at least >12 ppm. In contrast to that trend, a

small resonance at around 6.9 ppm is present in our sample

(Figure 5C inset). This signal is not observed in the sample

from the reaction of MSN and HBpin (Figure S6b in SI). The

signal at ∼7 ppm could not be unambiguously identiﬁed above

the noise. However, ²H DPMAS spectrum of ZrD/Bpin@MSN

zirconium-bound hydrogen chemical shifts in Cp*₂ZrH_2,

from the reaction of DBpin and Zr(NMe₂)_n@MSN contains

signals at 7.3, 4.0, 2.5 (as a shoulder), and 1.5 ppm

corresponding to deuterium incorporation into ZrH, NMe₂,

and Bpin groups (Figure S9). On the basis of these facile H/D

exchange reactions, this resonance is assigned as a zirconium

hydride. Spin counting experiments indicate that the ZrH

loading is ca. 0.5 mmol per gram, and thus surface ZrH sites

(Cp*₂ZrH)₂O, and Cp*₂ZrH(NH₂) (Cp* = C₅Me₅) were

reported to be 7.46, 5.5, and 4.82 ppm, respectively,^61,62and

even further upﬁeld ZrH resonances were reported for

Cp*₂ZrH(NH₂BH₃).⁶³That is, association of pinacolborane

or borate groups or amide moieties with a surface-bonded

zirconium hydride might result in upﬁeld chemical shifts with

respect to (SiO)₃ZrH and (SiO)₂ZrH₂. To test for the

account for >50% of the zirconium in the sample.

presence of a surface-supported zirconium hydride that is

distinct from (SiO)₃ZrH and (SiO)₂ZrH₂and assign the

¹H NMR signal at 6.9 ppm, the HBpin-treated solid was

allowed to react with D₂gas in benzene (eq 4).

The infrared spectrum of the ZrH/Bpin@MSN solid product

further supported this assignment (Figure 7B) on the basis of a

band centered at 1592 cm⁻¹that we assigned to a ν_ZrH. The IR

spectrum of MSN treated with HBpin does not contain signals

in this region (Figure 7A). Previously, a signal at 1638 cm⁻¹

was assigned to the ν_ZrHin (SiO)₃ZrH.⁴¹Importantly, the

Upon treatment with 1 atm of D₂, the signal at 6.9 ppm

diminished 50%, and after 3 cycles with 1 atm of D₂, the signal

7407

DOI: 10.1021/acscatal.5b01671

ACS Catal. 2015, 5, 7399−7414

ACS Catalysis

Research Article

characterization of this material as containing a zirconium

hydride, including the expected H/D exchange reactivity.⁶

Finally, additional experiments were performed to account

for the small amount of NMe₂groups in the reaction of HBpin

and Zr(NMe₂)_n@MSN (which produced only 2.5 mmol of

Me₂NBpin vs 2.7 mmol per gram of Zr(NMe₂)_n@MSN).

Combustion analysis of ZrH/Bpin@MSN revealed 0.4 0.05

mmol of nitrogen per gram of material. This value is

signiﬁcantly reduced in comparison to the Zr(NMe₂)_n@MSN

starting sample. Therefore, a ¹⁵N−¹⁵H idHetcor experiment

was used to probe the identity of the nitrogen species (Figure

8). There is a correlation from a ¹⁵N NMR signal at −355 ppm

Figure 6. ¹H DPMAS spectra of Zr(NMe₂)_n@MSN + 10 HBpin, (A)

as synthesized, (B) 1× of D₂exchange, (C) 3× of D₂exchange. The

spectra are normalized to the sample amount and show the absolute

intensities. The spectra were obtained using ν_R= 40 kHz, ν_RF(¹H) =

125 kHz, and τ_RD= 20 s.

1592 cm⁻¹peak was not detected in the IR spectrum of the

ZrH/Bpin@MSN exposed to D₂, and this change is taken as

evidence for the formation of ZrD/Bpin@MSN (Figure 7C).

Unfortunately, the expected location of a ν_ZrDat 1125 cm⁻¹

overlaps with the silica absorptions, and that signal could not be

detected. However, the signal at 1592 cm⁻¹reappears upon

addition of H₂to ZrD/Bpin@MSN (Figure 7D). Interestingly,

new broad signals at ∼2395 cm⁻¹appear in the sample treated

with D₂. Similar bands were observed in the spectrum resulting

from treatment of Zr(NMe₂)_n@MSN with DBpin (Figure 7E);

in that IR spectrum, the signal at 1592 cm⁻¹was also not

detected. These lower energy bands at ∼2395 cm⁻¹may be

attributed to H/D exchange reactions catalyzed by a surface

zirconium hydride and correspond to signals of deuterium-

exchanged pinacol and amido methyls. These observations are

consistent with the ¹H SSNMR results and support the

Figure 8. 2D ¹⁵N−¹H idHetcor spectrum of ¹⁵N-enriched ZrH/

Bpin@MSN. The spectrum was obtained at 14.1 T, using ν_R= 34

kHz; ν_RF(¹H) = 100 kHz during 90° pulse and 91 kHz during CP, and

10 kHz during SPINAL-64 ¹H decoupling; ν_RF(¹⁵N) = 62 kHz during

90° pulse and 57 kHz during CP pulse, and 10 kHz during SPINAL-64

¹⁵N decoupling; 128 rows with Δt₁= 30 μs; 64 scans per row, and

STATES-TPPI acquisition with τ_RD= 2 s.

1

to a H NMR resonance at 2.4 ppm. On the basis of the

similarity of this chemical shift to Me₂NBpin and the ¹¹B NMR

signal at 24 ppm, we attribute the residual ¹⁵N SSNMR signal

Figure 7. Diﬀuse reﬂectance IR spectra of (A) Bpin@MSN, (B) ZrH/Bpin@MSN, (C) ZrH/Bpin@MSN + D₂, (D) ZrD/Bpin@MSN + H₂, and

(E) Zr(NMe₂)_n@MSN + DBpin.

7408

DOI: 10.1021/acscatal.5b01671

ACS Catal. 2015, 5, 7399−7414

ACS Catalysis

Research Article

Table 3. Characterization Experiments and Conclusions Regarding the Nature of ZrH/Bpin@MSN

experiments

observations

interpretations

reaction stoichiometry: 2.5 mmol Me₂NBpin formed/g Zr(NMe₂)_n@MSN

all but ca. 0.2 mmol NMe₂/g are desorbed from the material

Zr(NMe₂)_n@MSN +

HBpin

reaction side products:

Zr(NMe₂)_n@MSN +

HBpin

small amount of H₂formed

large amount of H₂formed

few reactive silanols or NH groups present in Zr(NMe₂)_n@MSN

MSN + HBpin

ICP-OES:

the Zr:B ratio is ∼ 1:1.

Zr(NMe₂)_n@MSN +

HBpin:

0.89 mmol Zr/g material

0.86 mmol B/g material

1.33 mmol B/g material

ν_ZrHobserved at 1592 cm⁻¹

HBpin treatment does not leach Zr from MSN

MSN + HBpin:

IR

Bpin is grafted to the material

zirconium hydride formed using HBpin that is distinct from (

SiO)₃ZrH

reaction with D₂, then band at 1592 cm⁻¹disappears upon D₂addition, then reappears

H₂

exchangable zirconium hydride

upon H₂addition

¹H SSNMR:

treatment with D₂

¹¹B SSNMR:

MSN + HBpin

δ_ZrHat 6.9 ppm, 0.5 mmol H/g

signal disappears upon D₂addition

the zirconium hydride surface species is distinct from (

SiO)₃ZrH

the SiOBpin chemical environment is inﬂuenced by surface Zr

species

1

δ ≈ 19 ppm; δ_CS≈ 21 ppm; H−¹¹B distance ∼3.4 Å

1

Zr(NMe₂)_n@MSN +

HBpin

¹¹B NMR spin count

¹H−¹⁵N Hetcor

δ ≈ 18 ppm; δ_CS≈ 21 ppm; H−¹¹B distance ∼2.1 Å

0.9 mmol B/g

−355 ppm ¹⁵N signal correlates with 7.9 ppm H signal

1

residual NMe₂groups unreactive due to zirconium-coordination

and H-bonding

partly to surface-absorbed Me₂NBpin. Furthermore, there is

Further support for this idea is provided by related

homogeneous catalysis. The compound Zr(NMe₂)₄is an

eﬀective catalyst for this carbonyl hydroboration reaction, as

is {PhB(Ox^Me2)₂C₅H₄}Zr(NMe₂)₂.⁴⁶In the presence of 5 mol

% of either soluble complex, quantitative conversion of PhCHO

to PhCH₂OBpin is observed after 30 min in benzene-d₆at 60

°C. The faster conversion obtained with homogeneous versus

heterogeneous catalysts may result from the eﬀect of diﬀusion,

a lower percentage of active sites or slower site activation in the

supported catalyst, or simply the eﬀect of silica as a ligand for

zirconium in this catalysis.

also a correlation between the ¹⁵N SSNMR signal at −355 ppm

1

and a signal in the H NMR dimension at 7.9 ppm. These

1

chemical shifts, as well as the H−¹⁵N correlation, suggest that

some of the remaining surface nitrogen is present as Zr−

NHMe₂.

The evidence supporting the identity of ZrH/Bpin@MSN as

a zirconium hydride formed from reaction of Zr(NMe₂)_n@

MSN and HBpin is given in Table 3, and a description of the

spectroscopy and structural assignment is summarized in the

conclusion.

Catalytic Hydroboration of Carbonyls. On the basis of

the facile reaction of Zr(NMe₂)_n@MSN and HBpin, this

material was investigated as a catalyst for the hydroboration of

carbonyl compounds with pinacolborane. Initially, benzalde-

hyde was used as a test substrate to compare the reactivity of

supported zirconium with possible background reactions and

homogeneous analogues. With grafted amidozirconium as the

precatalyst (5 mol %, based on ICP-OES-determined zirconium

loading), quantitative conversion of benzaldehyde to its

pinacolborane ester is accomplished at room temperature

after 2 h in benzene-d₆, as determined by ¹H NMR

spectroscopy (eq 5). Control experiments, in which PhCHO

The hydroboration using Zr(NMe₂)_n@MSN as the catalyst is

selective in the presence of a number of functional groups, as

determined by conversion of substituted benzaldehydes shown

in Table 5. Aldehyde substrates containing ethers (p-methoxy-

benzaldehyde and furfural), nitro groups (p-nitrobenzalde-

hyde), halides (p-chlorobenzaldehyde), alkyl substitution (p-

tolualdehyde), an aliphatic aldehyde (cyclohexanecarboxalde-

hyde), and ferrocene substitution (ferrocene-2-carboxaldehyde)

are readily reduced, although para-chlorobenzaldehyde re-

quired 5× greater reaction time than benzaldehyde. Para-

substitution by formyl or pinacolborane ester groups does not

impact the reacting moiety as assessed by the hydroboration of

p-phthaldialdehyde which gives 1,4-bis(pinacolborane ester)-

benzene. Equivalent amounts of the solvent, reactants, and

catalysts are used in each experiment for consistency and

straightforward comparisons between substrates.

The hydroboration of ketones is also eﬃciently catalyzed by

Zr(NMe₂)_n@MSN. A background screen of catalyst-free

conditions, calcined MSN, or Bpin@MSN as catalysts for the

addition of acetophenone and pinacolborane only returned

unreacted acetophenone, even with heating to 100 °C (see

Table 6). A loading of 5 mol % Zr(NMe₂)_n@MSN catalyzes

quantitative formation of 1-phenylethoxyborane ester after 24 h

at room temperature in benzene. However, 60 °C appears to be

a generally appropriate temperature for convenient rates of

conversion. As in the aldehyde hydroboration examples,

and HBpin are mixed in benzene at room temperature with or

without MSN, give only trace quantities of product (Table 4).

In addition, the minimal conversion of PhCHO and HBpin

occur to the same extent with the zirconium-free material

Bpin@MSN, obtained from the reaction of MSN and HBpin, as

the slow, uncatalyzed background reaction. Thus, the zirconium

sites present in the Zr(NMe₂)_n@MSN material are responsible

for catalytic activity.

7409

DOI: 10.1021/acscatal.5b01671

ACS Catal. 2015, 5, 7399−7414

ACS Catalysis

Research Article

a

Table 4. Catalytic Hydroboration of Benzaldehyde and Acetophenone with Pinacolborane

a

b

5 mol % catalyst in benzene-d₆using 1.3 equiv of HBpin. Obtained by integration of product signal in comparison to Si(SiMe₃)₄as an internal

standard.

reactions of acetophenone and HBpin with 5 mol %

Zr(NMe₂)₄or {PhB(Ox^Me2)₂C₅H₄}Zr(NMe₂)₂(under homo-

geneous conditions) is accomplished in shorter times than with

5 mol % Zr(NMe₂)_n@MSN.

order to test for this possibility, the Zr(NMe₂)_n@MSN was

reacted with excess HBpin and the MSN product was analyzed.

Low-angle powder XRD (Figure S1 in SI) and TEM

measurements (Figure 1D,E) indicated that the pore structure

and particle morphology were not aﬀected by the pinacolbor-

ane. In addition, EDX measurements (Figure 1F) show that the

well-distributed zirconium remains unchanged after treatment

with HBpin. Thus, the zirconium-surface interaction and the

silica wall structure is maintained in the presence of HBpin.

A number of additional experiments also were performed to

test for zirconium leaching. First, the catalytic material was

isolated by ﬁltration, washed with benzene, dried under

vacuum, and reused in hydroboration of benzaldehyde or

acetophenone. This sequence was performed eight times with

both PhCHO and PhC(O)Me as substrates without apparent

loss of catalytic activity (Figure 9). The reactions were

monitored during the conversion, verifying that ∼2 h are

required for full conversion in the ﬁrst and eighth cycles.

Moreover, plots of acetophenone concentration versus time

roughly follow exponential decay, with the observed pseudo

ﬁrst-order rate constants after 1, 4, and 8 recycles being 4 ×

10⁻⁴s⁻¹, 4 × 10⁻⁴s⁻¹, and 3 × 10⁻⁴s⁻¹. Thus, the rates of

catalysis are not signiﬁcantly diminished with repeated catalyst

recycling. In the ﬁrst cycle with Zr(NMe₂)_n@MSN as the

precatalyst, the Me₂NBpin byproduct of catalyst activation is

present in the crude RR′HCO−Bpin product. This substance

was not observed during subsequent cycles, and pure boronate

ester product is obtained after ﬁltration and evaporation of

volatile materials.

As in the aldehyde hydroboration, aliphatic substituents with

α-hydrogen are reduced without production of pinacolborane

enolate ester side products that might form through substrate

deprotonation (Table 6). Linear aliphatic ketones are reduced

more rapidly than cyclic aliphatic or aryl-substituted ketones.

Ketones containing nitroarene or triﬂuoromethyl groups are

readily reduced without aﬀecting the functionality. In addition,

α,β-unsaturated ketones are reduced selectively at the carbonyl,

leaving the carbon−carbon double-bond intact. Benzophenone

and ﬂuorenone are also reduced in good yield, with ﬂuorenone

showing faster conversion under equivalent conditions.

A possible intermediate in the zirconium-catalyzed carbonyl

hydroboration is a zirconium alkoxy moiety of the type [Zr]−

OCHRR′. The reaction of such an intermediate with

pinacolborane to form a B−O bond is not necessarily

straightforward given the oxophilicity of zirconium. We should

note, however, that our recently proposed boron-centered

zwitterionic mechanism for a magnesium-catalyzed hydro-

boration of esters avoids the magnesium alkoxide intermedi-

ate.²⁴A related mechanism could bypass the Zr−O bond in the

current catalysis. Despite this possibility, magnesium alkoxides

and HBpin react to give pinacolborane esters. Moreover, the

surface-supported zirconium amide and HBpin react to give

Me₂NBpin.

Thus, the intermediacy of [Zr]OCHRR′ is not ruled out, and

there are (at least) two types of Zr−O bonds present in a

possible (SiO)_nX_3−nZr−OCHRR′ intermediate, a siloxide-

zirconium bond and an alkoxide-zirconium bond. Both moieties

might be capable of reaction with HBpin, with the reaction of

SiOZr bonds potentially resulting in catalyst leaching. In

Second, the zirconium loading on MSN after catalysis, as

determined by ICP-OES, is identical within error to the

loadings after grafting of Zr(NMe₂)₄on calcined MSN and

after heating at 60 °C in benzene. The same weight % Zr is

obtained for the ZrH/Bpin@MSN material, as obtained from

7410

DOI: 10.1021/acscatal.5b01671

ACS Catal. 2015, 5, 7399−7414

ACS Catalysis

Research Article

a

Table 5. Zr(NMe₂)_n@MSN-Catalyzed Hydroboration of Aldehydes with Pinacolborane

a

b

All the reactions are carried with 5 mol % Zr(NMe₂)_n@MSN in benzene-d₆at room temperature using 1.3 equiv of HBpin. Obtained by

c

d

integration of RCH₂OBpin signal against Si(SiMe₃)₄as an internal standard. Isolated yields for RCH₂OBpin product. Isolated yield of RCH₂OH

product after hydrolysis with NaOH. 2 equiv of HBpin was used.

e

the reaction of HBpin and Zr(NMe₂)_n@MSN and after

catalytic hydroboration reactions (Table 1). In addition, a

catalytic reaction mixture was ﬁltered after 50% conversion to

give a mixture of PhCOMe, HBpin, and PhMeHCO-Bpin, and

the soluble portion of the reaction mixture was heated at 60 °C.

The ratio of starting material and product in this separated

solution-phase portion is invariant over 1 h (i.e., no further

conversion), whereas full conversion to PhMeHCO−Bpin is

observed in an unﬁltered parallel experiment. Finally, the

supernatant was evaporated after a catalytic reaction, and only a

trace amount of zirconium (0.001 mM) was detected by ICP-

OES. These experiments reinforce the robust nature of the

supported zirconium catalyst and the supported nature of the

hydroboration catalyst.

CONCLUSION

■

The reaction of Zr(NMe₂)₄and calcined mesoporous silica

provides Zr(NMe₂)_n@MSN, a material containing zirconium

sites with a Zr:NMe₂ratio of ∼1:2.7. Detailed SSNMR studies,

particularly indirectly detected ¹⁵N Hetcor experiments and

DNP-enhanced CPMAS ¹⁵N NMR spectra, reveal that

zirconium is primarily bonded to dimethylamide groups, with

a small amount of coordinated dimethylamine. These data,

together with quantitative ¹³C SSNMR and elemental analysis,

characterize the surface zirconium species as containing three

7411

DOI: 10.1021/acscatal.5b01671

ACS Catal. 2015, 5, 7399−7414

ACS Catalysis

Research Article

a

Table 6. Zr(NMe₂)_n@MSN-Catalyzed Hydroboration of Ketones with Pinacolborane

a

b

All the reactions are carried with 5 mol % catalyst in benzene-d₆at 60 °C using 1.3 equiv of HBpin. Obtained by integration of R₂CHOBpin signal

c

d

against Si(SiMe₃)₄as an internal standard. Isolated yields for R₂CHOBpin. Isolated yield for R₂CHOH.

sites: primarily (more than 70%) coordinated by three

nitrogen-containing ligands, with the remaining sites (up to

30%) as the dipodal (SiO)₂Zr(NMe₂)₂. Moreover, the

former sites are a mixture of mainly monopodal SiOZr-

(NMe₂)₃(more than 90%) with the small remaining amount as

diamido amine (SiO)₂Zr(NMe₂)₂(NHMe₂). The basis for

7412

DOI: 10.1021/acscatal.5b01671

ACS Catal. 2015, 5, 7399−7414

ACS Catalysis

Research Article

MSN are active in carbonyl hydroboration even after exposure

of the catalytic materials to air. For example, reaction of

Zr(NMe₂)_n@MSN and air produces detectable amounts of

HNMe₂, but the addition reaction of HBpin and acetophenone

is readily catalyzed by the hydrolyzed material under the

conditions of Table 4. Thus, this MSN-supported early

transition metal system is a capable, robust catalyst for the

reduction of oxygenated organic compounds.

Although the catalytic reduction of oxygenates by early

transition metal sites has previously been demonstrated with

homogeneous catalysis, zirconium hydrides are typically

associated with extreme sensitivity to air and moisture. This

list of highly sensitive species includes silica surface-supported

zirconium hydrides, which have previously shown high

reactivity toward inert substrates such as methane.^1,3,6,40

Although the apparent rate of hydroboration with surface-

supported Zr(NMe₂)_n@MSN is lower than the homogeneous

zirconium catalysts, the accessibility of a catalytically active

species even after air exposure and the recyclability of the

surface-supported catalyst provide appealing advantages for the

heterogeneous system. We are currently exploring other

hydride sources to access surface-supported hydrides for new

catalytic applications.

Figure 9. Catalyst recycling experiments. Quantitative conversion and

>99% selectivity for acetophenone hydroboration is observed after

reisolation and reapplication of the catalyst for at least eight cycles. In

situ monitoring of the catalytic reactions indicated that ∼2 h is

required for full conversion in each recycle experiment.

ASSOCIATED CONTENT

* Supporting Information

■

S

The Supporting Information is available free of charge on the

ACS Publications website at DOI: 10.1021/acscatal.5b01671.

1

this conclusion is the reactivity of Zr(NMe₂)_n@MSN with

HBpin. Around 85−90% of the surface NMe₂groups react with

HBpin to give Me₂NBpin, and the remaining nitrogen groups

are likely present as either zirconium-coordinated and hydro-

gen-bond species or physisorbed Me₂NBpin. This assignment is

supported by the characteristic ¹⁵N SSNMR chemical shift

(−355 ppm), the nitrogen correlation with a broad downﬁeld

hydrogen signal in ¹⁵N−¹H Hetcor experiments, and the high

reactivity of both dimethylamine and dimethyl amidozirconium

groups toward pinacolborane. The latter observations suggest

that the unreactive NMe₂are in chemically distinct environ-

ments from free HNMe₂and ZrNMe₂moieties, and

coordination to zirconium and the involvement of the NMe₂

groups in hydrogen-bonding to the silica surface may explain

their inert nature.

Powder XRD patterns, H NMR spectra of Zr(NMe₂)₄

grafting and HBpin reactions for quantiﬁcation, and

1

catalysis products, and solid-state H and ¹¹B DPMAS

NMR spectra (PDF)

AUTHOR INFORMATION

Corresponding Authors

*E-mail: sadow@iastate.edu.

*E-mail: mpruski@iastate.edu.

■

Present Address

^‡Exxon-Mobil

Notes

The authors declare no competing ﬁnancial interest.

The surface species formed in the reaction of Zr(NMe₂)_n@

MSN and HBpin, namely, ZrH/Bpin@MSN, was characterized

ACKNOWLEDGMENTS

■

1

by H and ¹¹B SSNMR and infrared spectroscopies. In this

This research was supported by the U.S. Department of Energy,

Oﬃce of Basic Energy Sciences, Division of Chemical Sciences,

Geosciences, and Biosciences through the Ames Laboratory

(Contract No. DE-AC02-07CH11358). The authors thank

BASF for the generous donation of the P104 surfactant.

MSN system, the reaction of HBpin and silanol groups

provides SiOBpin, and the interaction of these groups and

surface zirconium species perturbs the chemical environment of

1

the surface Bpin groups. H NMR and infrared spectroscopies

reveal signals assigned to zirconium hydride, and these

assignments are supported by selective H/D exchange reactions

of zirconium hydride and D₂or zirconium deuteride with H₂.

Thus, the reaction of HBpin and Zr(NMe₂)_n@MSN provides a

zirconium hydride that shows the anticipated reactivity of such

a species in H/D exchange reactions.

The present work has demonstrated that surface-supported,

catalytically active zirconium hydrides are accessible from

amides and likely from metal alkoxides or hydroxides using

HBpin as the hydride source. That notion is advanced by

characterization data and the catalytic hydroboration studies,

which may involve the zirconium hydride and zirconium

alkoxides as catalytic intermediates. The zirconium sites on

REFERENCES

■

(1) Quignard, F.; Choplin, A.; Basset, J.-M. J. Chem. Soc., Chem.

Commun. 1991, 1589−1590.

(2) Zakharov, V. A.; Dudchenko, V. K.; Paukshtis, E. A.; Karakchiev,

L. G.; Yermakov, Y. I. J. Mol. Catal. 1977, 2, 421−435.

(3) Thieuleux, C.; Quadrelli, E. A.; Basset, J.-M.; Dobler, J.; Sauer, J.

Chem. Commun. 2004, 1729−1731.

(4) Zakharov, V. A.; Yermakov, Y. I. Catal. Rev.: Sci. Eng. 1979, 19,

67−103.

(5) Yermakovt, Y. I.; Ryndin, Y. A.; Alekseev, O. S.; Kochubey, D. I.;

Shmachkov, V. A.; Gergert, N. I. J. Mol. Catal. 1989, 49, 121−132.

(6) Casty, G. L.; Matturro, M. G.; Myers, G. R.; Reynolds, R. P.; Hall,

R. B. Organometallics 2001, 20, 2246−2249.

7413

DOI: 10.1021/acscatal.5b01671

ACS Catal. 2015, 5, 7399−7414

ACS Catalysis

Research Article

(7) Basset, J.-M.; Coperet, C.; Soulivong, D.; Taoufik, M.; Cazat, J. T.

Acc. Chem. Res. 2009, 43, 323−334.

́

(8) Coperet, C.; Chabanas, M.; Petroff Saint-Arroman, R.; Basset, J.-

(44) Kandel, K.; Frederickson, C.; Smith, E. A.; Lee, Y.-J.; Slowing, I.

I. ACS Catal. 2013, 3, 2750−2758.

(45) Diamond, G. M.; Jordan, R. F.; Petersen, J. L. J. Am. Chem. Soc.

1996, 118, 8024−8033.

M. Angew. Chem., Int. Ed. 2003, 42, 156−181.

(9) Lecuyer, C.; Quignard, F.; Choplin, A.; Olivier, D.; Basset, J.-M.

Angew. Chem., Int. Ed. Engl. 1991, 30, 1660−1661.

(10) Quignard, F.; Lecuyer, C.; Choplin, A.; Olivier, D.; Basset, J.-M.

(46) Manna, K.; Ellern, A.; Sadow, A. D. Chem. Commun. 2010, 46,

339−341.

́

(47) Gutekunst, G.; Brook, A. G. J. Organomet. Chem. 1982, 225, 1−

́

3.

J. Mol. Catal. 1992, 74, 353−363.

(48) Gilman, H.; Smith, C. L. J. Organomet. Chem. 1967, 8, 245−253.

(49) Wu, J. Y.; Moreau, B.; Ritter, T. J. Am. Chem. Soc. 2009, 131,

12915−12917.

(11) Carter, M. B.; Schiott, B.; Gutierrez, A.; Buchwald, S. L. J. Am.

Chem. Soc. 1994, 116, 11667−11670.

(12) Fu, P.-F.; Brard, L.; Li, Y.; Marks, T. J. J. Am. Chem. Soc. 1995,

117, 7157−7168.

(50) Lesage, A.; Lelli, M.; Gajan, D.; Caporini, M. A.; Vitzthum, V.;

́

Mieville, P.; Alauzun, J.; Roussey, A.; Thieuleux, C.; Mehdi, A.;

(13) Gountchev, T. I.; Tilley, T. D. Organometallics 1999, 18, 5661−

5667.

Bodenhausen, G.; Coperet, C.; Emsley, L. J. Am. Chem. Soc. 2010, 132,

15459−15461.

(14) Molander, G. A.; Julius, M. J. Org. Chem. 1992, 57, 6347−6351.

(15) Molander, G. A.; Retsch, W. H. Organometallics 1995, 14,

4570−4575.

(16) Takahashi, T.; Hasegawa, M.; Suzuki, N.; Saburi, M.; Rousset,

C. J.; Fanwick, P. E.; Negishi, E. J. Am. Chem. Soc. 1991, 113, 8564−

8566.

(17) Xin, S. X.; Harrod, J. F. Can. J. Chem. 1995, 73, 999−1002.

(18) Jeske, G.; Lauke, H.; Mauermann, H.; Schumann, H.; Marks, T.

J. J. Am. Chem. Soc. 1985, 107, 8111−8118.

(51) Zagdoun, A.; Casano, G.; Ouari, O.; Schwarzwalder, M.; Rossini,

̈

́

A. J.; Aussenac, F.; Yulikov, M.; Jeschke, G.; Coperet, C.; Lesage, A.;

Tordo, P.; Emsley, L. J. Am. Chem. Soc. 2013, 135, 12790−12797.

(52) Chisholm, M. H.; Hammond, C. E.; Huffman, J. C. Polyhedron

1988, 7, 2515−2520.

(53) Fujdala, K. L.; Tilley, T. D. J. Am. Chem. Soc. 2001, 123, 10133−

10134.

(54) Quignard, F.; Lecuyer, C.; Bougault, C.; Lefebvre, F.; Choplin,

A.; Olivier, D.; Basset, J. M. Inorg. Chem. 1992, 31, 928−930.

(55) Troullier, N.; Martins, J. L. Phys. Rev. B: Condens. Matter Mater.

Phys. 1991, 43, 1993−2006.

(19) Tilley, T. D. Acc. Chem. Res. 1993, 26, 22−29.

(20) Gauvin, F.; Harrod, J. F.; Woo, H. G. Adv. Organomet. Chem.

1998, 42, 363−405.

(56) Amoureux, J. P.; Pruski, M. Mol. Phys. 2002, 100, 1595−1613.

(57) Althaus, S. M.; Mao, K.; Stringer, J. A.; Kobayashi, T.; Pruski, M.

Solid State Nucl. Magn. Reson. 2014, 57−58, 17−21.

(21) Waterman, R. Organometallics 2013, 32, 7249−7263.

(22) Arrowsmith, M.; Hadlington, T. J.; Hill, M. S.; Kociok-Kohn, G.

Chem. Commun. 2012, 48, 4567−4569.

(58) Sole,

11355.

́ ́

C.; Fernandez, E. Angew. Chem., Int. Ed. 2013, 52, 11351−

(23) Harrison, K. N.; Marks, T. J. J. Am. Chem. Soc. 1992, 114, 9220−

9221.

(24) Mukherjee, D.; Ellern, A.; Sadow, A. D. Chem. Sci. 2014, 5,

(59) Medek, A.; Harwood, J. S.; Frydman, L. J. Am. Chem. Soc. 1995,

117, 12779−12787.

(60) Fernandez, C.; Pruski, M. Solid State NMR; Chan, J. C. C., Ed.;

Springer: Berlin, 2012; Vol. 306, pp 119−188.

(61) Manriquez, J. M.; McAlister, D. R.; Sanner, R. D.; Bercaw, J. E. J.

Am. Chem. Soc. 1976, 98, 6733−6735.

(62) Hillhouse, G. L.; Bercaw, J. E. J. Am. Chem. Soc. 1984, 106,

5472−5478.

(63) Almqvist, F.; Torstensson, L.; Gudmundsson, A.; Frejd, T.

Angew. Chem., Int. Ed. Engl. 1997, 36, 376−377.

959−964.

(25) Arrowsmith, M.; Hill, M. S.; Hadlington, T.; Kociok-Kohn, G.;

Weetman, C. Organometallics 2011, 30, 5556−5559.

(26) Anker, M. D.; Arrowsmith, M.; Bellham, P.; Hill, M. S.; Kociok-

Kohn, G.; Liptrot, D. J.; Mahon, M. F.; Weetman, C. Chem. Sci. 2014,

5, 2826−2830.

(27) Oluyadi, A. A.; Ma, S.; Muhoro, C. N. Organometallics 2012, 32,

70−78.

(28) Khalimon, A. Y.; Farha, P.; Kuzmina, L. G.; Nikonov, G. I.

Chem. Commun. 2012, 48, 455−457.

(29) Koren-Selfridge, L.; Query, I. P.; Hanson, J. A.; Isley, N. A.;

Guzei, I. A.; Clark, T. B. Organometallics 2010, 29, 3896−3900.

(30) Evans, D. A.; Fu, G. C. J. Org. Chem. 1990, 55, 5678−5680.

(31) Hadlington, T. J.; Hermann, M.; Frenking, G.; Jones, C. J. Am.

Chem. Soc. 2014, 136, 3028−3031.

(32) Verdaguer, X.; Lange, U. E. W.; Reding, M. T.; Buchwald, S. L. J.

Am. Chem. Soc. 1996, 118, 6784−6785.

(33) Willoughby, C. A.; Buchwald, S. L. J. Am. Chem. Soc. 1994, 116,

11703−11714.

(34) Willoughby, C. A.; Buchwald, S. L. J. Am. Chem. Soc. 1994, 116,

8952−8965.

(35) Yun, J.; Buchwald, S. L. J. Am. Chem. Soc. 1999, 121, 5640−

5644.

(36) Pereira, S.; Srebnik, M. Organometallics 1995, 14, 3127−3128.

(37) Eter, M. E.; Hamzaoui, B.; Abou-Hamad, E.; Pelletier, J. D. A.;

Basset, J.-M. Chem. Commun. 2013, 49, 4616−4618.

(38) Beaudoin, M.; Scott, S. L. Organometallics 2000, 20, 237−239.

(39) Hamzaoui, B.; Eter, M. E.; Abou-hamad, E.; Chen, Y.; Pelletier,

J. D. A.; Basset, J.-M. Chem. - Eur. J. 2015, 21, 4294−4299.

(40) Pasha, F. A.; Bendjeriou-Sedjerari, A.; Huang, K.-W.; Basset, J.-

M. Organometallics 2014, 33, 3320−3327.

(41) Rataboul, F.; Baudouin, A.; Thieuleux, C.; Veyre, L.; Coperet,

C.; Thivolle-Cazat, J.; Basset, J. M.; Lesage, A.; Emsley, L. J. Am. Chem.

Soc. 2004, 126, 12541−12550.

(42) Brown, H. C. Hydroboration; W. A. Benjamin: New York, 1962.

(43) Gaylord, N. G. Reduction with Complex Metal Hydrides;

Interscience: New York, 1956.

7414

DOI: 10.1021/acscatal.5b01671

ACS Catal. 2015, 5, 7399−7414

Article Doi

Mesoporous Silica-Supported Amidozirconium-Catalyzed Carbonyl Hydroboration

DOI: 10.1021/acscatal.5b01671

Source and publish data:

Authors:

Article abstract of DOI:10.1021/acscatal.5b01671

Full text of DOI:10.1021/acscatal.5b01671

Products guided by the article

R&D Labs maybe for 100-49-2

Relevant to this article

Hot Product