Article pubs.acs.org/Macromolecules
Polyacrylonitrile‑b‑poly(butyl acrylate) Block Copolymers as Precursors to Mesoporous Nitrogen-Doped Carbons: Synthesis and Nanostructure Maciej Kopeć,†,# Rui Yuan,†,# Eric Gottlieb,†,# Carlos M. R. Abreu,†,‡ Yang Song,†,§ Zongyu Wang,† Jorge F. J. Coelho,‡ Krzysztof Matyjaszewski,*,† and Tomasz Kowalewski*,† †
Department of Chemistry, Center for Macromolecular Engineering, Carnegie Mellon University, 4400 Fifth Avenue, Pittsburgh, Pennsylvania 15213, United States ‡ CEMMPRE, Department of Chemical Engineering, University of Coimbra, Polo II, Pinhal de Marrocos, 3030-790 Coimbra, Portugal § Collaborative Innovation Center of Advanced Nuclear Energy Technology, Institute of Nuclear and New Energy Technology, Tsinghua University, Beijing 100084, China S Supporting Information *
ABSTRACT: A series of polyacrylonitrile-block-poly(butyl acrylate) (PAN-b-PBA) copolymers were prepared by supplemental activator reducing agent atom transfer radical polymerization (SARA ATRP). These copolymers were then used as precursors to pyrolytic nanostructured carbons with the PAN block serving as a nitrogen-rich carbon precursors and the PBA block acting as a sacrificial porogen. The study revealed that while the size of mesopores can be controlled by the size of the porogenic block, the connectivity of pores diminishes with the decrease of the overall molecular weight of the precursor. This partial loss of mesopore connectivity was attributed to the weaker phase segregation between the blocks of shorter lengths inferred from the shape of small-angle X-ray scattering profiles and from the crystallinity of polyacrylonitrile phase.
■
in excess of 600 °C under anaerobic conditions, was realized through the use of a moderate temperature (200−300 °C) oxidative annealing step, commonly used in the synthesis of PAN-derived carbon fibers.21−23 Using this step allowed for successful preservation of the BCP morphology in the synthesis of nanostructured carbon films,8,24 particles,25 molecular brushes,26 and bulk mesoporous carbons.9,27 Other groups followed this approach and reported polyacrylonitrile-b-poly(methyl methacrylate) (PAN-b-PMMA),28−30 polyacrylonitrileb-polystyrene (PAN-b-PS),31−33 or poly(tert-butyl acrylate)-bpolyacrylonitrile (PtBA-b-PAN)34 BCPs as efficient carbon precursors. Furthermore, in our past studies of electrochemical applications of CTNCs, we have established that the desirable bicontinuous morphology can be obtained from BCPs containing ∼40 wt % of PAN.9,17 Use of these precursors routinely yields CTNCs with surface areas of ∼500 m2/g. One question that has not been addressed so far is the extent to which the pore size and surface area of CTNCs could be varied by using precursors with the same PAN/PBA ratio, but differing overall MW.
INTRODUCTION Mesoporous nitrogen-doped nanocarbons are currently receiving considerable attention due to a range of electrochemical properties, which make these materials of particular interest in sustainable energy-related applications.1−7 Over the past several years, we have developed a facile method to synthesize copolymer-templated nitrogen-enriched nanocarbons (CTNCs) through pyrolysis of nanostructured block copolymers (BCPs) containing polyacrylonitrile (PAN), namely poly(butyl acrylate)-block-polyacrylonitrile (PBA-b-PAN).8−10 The well-defined structure of those copolymers resulted from the use of atom transfer radical polymerization (ATRP) which provided precise control over their molecular weight (MW) and molecular weight distribution (MWD).11−16 Owing to the combination of high surface area and the presence of highly accessible electrochemically active nitrogen atoms, CTNCs were successfully used as supercapacitors,9,10 electrocatalysts for oxygen reduction reaction (ORR)17 and hydrogen evolution reaction (HER),18 CO2 sorbents,19 and cathodes in dyesensitized solar cells (DSSC).20 The high surface areas desirable in above-mentioned applications were achieved by the use of PBA-b-PAN BCPs of the composition assuring the formation of bicontinuous morphologies.9 The preservation of BCP nanostructure upon pyrolysis, which involves heating the material to temperatures © XXXX American Chemical Society
Received: December 11, 2016 Revised: March 6, 2017
A
DOI: 10.1021/acs.macromol.6b02678 Macromolecules XXXX, XXX, XXX−XXX
Article
Macromolecules
solution (1:1, v/v) was added to start polymerization at room temperature. The resulting block copolymer was precipitated into methanol/water (1:1, v/v), filtered, and dried under vacuum overnight. Preparation of Porous Carbons. Bulk carbons were prepared directly from precipitated powders. First, the BCP samples were stabilized by heating to 280 °C under air flow (150 mL/min) at a rate of 1 °C/min and kept at this temperature for 1 h. Stabilized samples were then pyrolyzed under nitrogen gas flow (150 mL/min) by heating to 800 °C at a rate of 10 °C/min and kept at this temperature for 30 min. Characterization. 1H NMR spectroscopy measurements were performed on a Bruker Avance 300 MHz spectrometer and used to determine the conversion of monomer in DMSO-d6 and Mn of resulting (co)polymers in DMSO-d6 for PAN homopolymer and DMF-d7 for block copolymers. The molecular weights (Mn) and molecular weight distributions (Mw/Mn) were determined by gel permeation chromatography (GPC). The GPC system used a Waters 515 HPLC pump and a Waters 2414 refractive index detector using Waters columns (Styrogel 102, 103, and 105 Å) with 10 mM LiBrcontaining DMF as the eluent at a flow rate of 1 mL/min at 50 °C using linear poly(ethylene oxide) (PEO) standards for PAN and PMMA for PAN-b-PBA copolymers.41 Differential scanning calorimetry (DSC) was performed on a TA Instruments Q20 and a Q2000 as well as on a Seiko DSC2100. Gas flow was kept at 20 mL/min of either purified air or N2. Sample sizes were between 1 and 5 mg. Thermogravimetric analysis (TGA) was performed on a TA Instruments Q50 with 60 mL/min flow rate of air or nitrogen. Brunauer−Emmett−Teller (BET) specific surface area measurements were carried out using a Micromeritics Gemini VII 2390 surface area analyzer with a VacPrep 061 degasser. Carbon samples were degassed at 300 °C and 20 mTorr vacuum for at least 8 h prior to measurement. The adsorption isotherms were fitted to the Barrett−Joyner−Halenda (BJH) model with the Kruk−Jaroniec−Sayari (KJS) correction to yield pore-size distributions. The surface area of micropores was estimated using the t-plot method with the KJS thickness correction. Transmission electron microscopy (TEM) (HT-7700, Hitachi Ltd., Tokyo, Japan) was conducted at an accelerating voltage of 120 kV. Small-angle X-ray scattering (SAXS) was performed using a Rigaku SMAX3000 instrument equipped with a 1.77 m sample-to-detector distance, a sealed microfocus source (Cu Kα, λ = 0.154 18 nm), and a wire-array detector (1024 by 1024, pixel size 0.1 mm). SAXS was collected with powders packed into a 5 mm thick sample holder, with Scotch tape sealing the powders. Wide angle X-ray scattering patterns were collected at the D1 beamline at Cornell High Energy Synchrotron Source. All BCP samples analyzed were annealed prior to scattering measurements by overnight heating at 160 °C under vacuum. Since all samples were isotropic, 2D patterns (Figure S5) were converted into 1D radial decay profiles in q space using custom procedures written in Mathematica (Wolfram Research, Inc.).
In the present study, we were able to address this issue by resorting to a newly developed synthetic approach via supplemental activator reducing agent (SARA) ATRP.13,35−37 Since the controlled synthesis of PAN-b-PBA copolymers is challenging due to different activities of acrylonitrile (AN) and butyl acrylate (BA) and low solubility of PAN, the first part of this report is dedicated to the detailed description of the development of appropriate synthetic conditions. The SARA ATRP enabled preparation of BCPs with precisely controlled block lengths and low dispersities (Mw/Mn < 1.2) at room temperature. Additionally, SARA ATRP allowed for reducing the copper catalyst concentration to only 50 ppm vs. monomer from previously used >1000 ppm.9,15 This is of particular importance for potential applications of CTNCs, as any residual metal may obscure the origin of catalytic activity in N-doped carbons.38 The structural studies described in the second part focused on exploring the correlation between the BCP and CTNC nanostructure using the combination of small-angle X-ray scattering (SAXS), transmission electron microscopy (TEM), and nitrogen adsorption. The impact of the overall molecular weight of BCPs on CTNC morphology was studied using a series of BCPs with the overall content demonstrated previously to yield bicontinuous morphologies. We show that while the size of mesopores in CTNCs can be indeed systematically varied by total MW of a BCP precursor, the accessible surface area is limited by the weak segregation in lowMW BCPs. In addition, thermal analysis was used to investigate the relationship between reactions occurring in both blocks during transformation into nanocarbon and wide-angle X-ray scattering (WAXS) revealed the dependence of the crystallinity of PAN blocks on the MW of the precursor.
■
EXPERIMENTAL SECTION
Materials. Acrylonitrile (AN, Sigma-Aldrich, >99%) and n-butyl acrylate (BA, Sigma-Aldrich, >99%) were purified by passing over a column of basic alumina to remove the inhibitor. 2-Bromopropionitrile (BPN, Sigma-Aldrich, 97%), 2,2′-bipyridine (bpy, Sigma-Aldrich, >99%), N,N,N′,N″,N″-pentamethyldiethylenetriamine (PMDETA, Sigma-Aldrich, 99%), copper(II) bromide (CuBr2, Acros Organics, >99%), copper wire (diameter 1.0 mm, 99.9+%, Aldrich), dimethylformamide (DMF, Fisher, 99.9%), dimethyl sulfoxide (DMSO, Fisher, 99.9%), methanol (Fisher, 99.9%), and diethyl ether (Fischer, 99%) were used as received. Tris(2-pyridylmethyl)amine (TPMA) was synthesized according to a published procedure.39,40 Polymerization of AN by SARA ATRP. In a typical procedure, copper wire (3 cm × 1 mm) was placed in a 10 mL Schlenk flask. Monomer (AN) was bubbled with nitrogen for 20 min, and 3.0 mL (45.3 mmol) was injected into the reactor under nitrogen. Initiator (BPN, 31.3 mg, 0.227 mmol) was then injected into a mixture of CuBr2 (0.51 mg, 0.002 mmol), bpy (1.06 mg, 0.007 mmol), DMF (0.5 mL, NMR standard), and DMSO (4.5 mL), previously bubbled with nitrogen for 20 min. The resulting mixture was transferred to the Schlenk flask under nitrogen. The reaction was conducted at room temperature. Samples of the reaction mixture were collected periodically during the polymerization by using an airtight syringe while purging the side arm of the Schlenk flask with nitrogen. Chain Extension of PAN-Br with BA by SARA ATRP. In a typical procedure, 8.23 g (1.75 mmol, 1 equiv) of PAN-Br macroinitiator (Mn,NMR = 4700, Mw/Mn = 1.25) was dissolved in 50 mL of DMF. Then, the solution was added to a 100 mL Schlenk flask containing 1.37 mg of CuBr2 (0.0035 mmol, 0.01 equiv) and 5.34 mg of TPMA (0.018 mmol, 0.0105 equiv) and bubbled with nitrogen for 30 min. 17.5 mL of deoxygenated BA (122 mmol, 70 equiv) was then carefully added under vigorous stirring to prevent PAN precipitation. Cu wire (14 cm × 1 mm), previously cleaned with HCl:MeOH
■
RESULTS AND DISCUSSION Synthesis of PAN-b-PBA Block Copolymers by SARA ATRP. Previous reports on synthesis of PAN block copolymers usually employed normal ATRP with halogen exchange42 to extend less active (meth)acrylates with more active AN.9,15,34 In order to prepare PAN-b-PBA BCPs without the halogen exchange and thus reduce the catalyst concentration, PAN macroinitiator had to be synthesized in the first step. Recently, we showed that initiators for continuous activator regeneration (ICAR) ATRP conducted at low loadings of a CuBr2/TPMA catalyst (1−50 ppm) yield well-defined PAN.41 However, ICAR ATRP requires the radical initiator which can also initiate additional polymer chains and increase the fraction of dead chains.43 Alternatively, SARA ATRP employs zerovalent Cu, typically in form of wire or powder, to reduce the CuIIX2/ Ligand(L) deactivator to CuIX/L activator species that subsequently activate the alkyl halide initiator. Additionally, B
DOI: 10.1021/acs.macromol.6b02678 Macromolecules XXXX, XXX, XXX−XXX
Article
Macromolecules
Figure 1. (a) First-order kinetic plots, (b) Mn,GPC and Mw/Mn evolution for SARA ATRP of AN with different catalysts. [AN]:[BPN]:[CuBr2]:[L] = 200:1:0.01:0.03, Cu wire (3 cm × 1 mm), [L] = TPMA, PMDETA or bpy; V0AN = 3 mL, AN:DMSO = 1:1.5 (v/v), rt. (c) GPC traces for [L] = bpy (Table 1, entry 4).
Table 1. SARA ATRP of AN with Different Catalytic Systemsa entry
a
ligand
1
TPMA
2 3 4
PMDETA bpy bpy
kpapp (h−1)
time (h)
conv (%)
Mn,theory
Mn,GPC
Mw/Mn
0.159 (20%) 0.084 0.079 0.051
6.8
41
4300
7800
1.44
7.3 8.0 36
43 46 83
4600 5100 9000
5300 5400 9000
1.24 1.24 1.19
[AN]:[BPN]:[CuBr2]:[ligand] = 200:1:0.01:0.03, Cu wire (3 cm × 1 mm); V0AN = 3 mL, AN:DMSO = 1:1.5 (v/v), rt.
Table 2. Summary of PAN-Br Chain Extension with BA via SARA ATRP PAN-b-PBA entry 1 2 3 4
DPPAN,
a
Mw/Mnb
70, 1.25 86, 1.25 122, 1.24 184, 1.21
[BA]:[PAN-Br]:[CuBr2]
conv (%)
composition
Mn, NMR
Mw/Mn
90:1:0.005 70:1:0.0035 125:1:0.00625 200:1:0.0075
42 80 60 76
(AN)70-b-(BA)40 (AN)86-b-(BA)54 (AN)122-b-(BA)79 (AN)184-b-(BA)124
8900 11600 16700 25800
1.18 1.24 1.19 1.29
Determined by 1H NMR. bDetermined by GPC. Entry 1: V0BA = 1.5 mL, BA:DMF 1:2 (v/v), Cu wire (1 cm × 1 mm), 5 h. Entry 2: V0BA = 17.5 mL, BA:DMF 1:3 (v/v), Cu wire (14 cm × 1 mm), 33 h. Entries 3 and 4: V0BA = 5.5 mL, BA:DMF 1:4 (v/v), Cu wire (5 cm × 1 mm), 24 and 96 h, respectively; [CuBr2]:[TPMA] = 1:3, rt. a
Cu0 acts as a supplemental activator by reacting directly with alkyl halide to yield a radical and CuIX/L.35−37 Three of the most widely used ATRP ligands, namely bpy, PMDETA, and TPMA, were examined at 50 ppm of a CuBr2/L catalyst loading (vs monomer) at room temperature in DMSO. BPN was used as the initiator due to its structural similarity with AN and high initiation efficiency. As shown in Figure 1, the reaction rate with a CuBr2/TPMA catalyst slowed down after 4 h and yielded PAN with Mn,GPC higher than Mn,theory. The rapid increase of the MW at conversions >20%, as well as broadening of the MWD to Mw/Mn = 1.44, indicated that radical termination occurred. Thus, less active catalyst complexes with PMDETA or bpy ligands were tested. Both catalysts provided linear first-order kinetics with almost equal polymerization rates as well as good agreement between Mn,GPC and Mn,theory (Table 1). Some deviations at low conversions, particularly visible for bpy, were caused by slow initiation with less active catalysts. Dispersity decreased for both PMDETA and bpy to values Mw/Mn < 1.25. These results differ from those obtained for ICAR ATRP, where the most active ligand (TPMA) provided the best control.41 In SARA ATRP, the ratelimiting steps, namely reduction of CuIIBr2/L by Cu(0) and supplemental activation, are faster than decomposition of a radical initiator in ICAR ATRP. Combined with high activity of AN and BPN initiator, CuBr2/TPMA led to fast activation, resulting in high concentration of radicals and increased rate of
termination. This points out that careful selection of reaction conditions in activator regeneration ATRP techniques is required, depending on the monomer, initiator, and reducing agent. CuBr2/bpy was then used as the catalyst to polymerize AN to high conversion (Table 1, entry 4). The reaction slowed down after reaching 50% conversion, and 83% conversion was obtained after 36 h. The final MW displayed good agreement with theoretical values and narrow MWD (Mw/Mn = 1.19). The structure and MW of isolated PAN were confirmed by 1H NMR (Figure S1). The synthesized PAN was then used as macroinitiator to prepare PAN-b-PBA block copolymers. As previously shown, chain extension in this system can be challenging, especially for macroinitiators with higher MW, since PAN is not soluble in BA.15 To overcome this issue, PAN was first dissolved in DMF overnight, and the BA:DMF ratio was maintained between 1:2 and 1:4 (v/v). Table 2 summarizes chain extension experiments of PAN macroinitiators of different MWs with BA. Reactions were conducted at room temperature with 50 ppm of CuBr2/ TPMA catalyst. A clean shift in GPC traces was observed, indicating formation of a block copolymer (Figure 2) with conversions from 42 to 80%. MW and composition of precipitated BCPs were determined by 1H NMR (Figure S2). All BCP samples showed excellent agreement between Mn,theory and Mn,NMR and low dispersity (Mw/Mn < 1.25), suggesting good control over the chain extension polymerization. A C
DOI: 10.1021/acs.macromol.6b02678 Macromolecules XXXX, XXX, XXX−XXX
Article
Macromolecules
maximum at 325−340 °C (Figure 3b) and an onset temperature of 250 °C corresponds to oxidative stabilization of PAN. Notably, heat dissipation over a broader temperature range is desirable as quick release of cyclization heat observed in nitrogen may lead to degradation of polymer morphology, an issue known in carbon fiber processing.23 Additionally, the presence of oxygen incorporated in the cross-linked PAN ladder-like structure leads to more efficiently stabilized polymer and facilitates further carbonization.44,45 A pronounced exotherm at 378−394 °C is attributed to the oxidative decomposition of PBA. Under air, the PBA block is removed at lower temperature and the process partially overlapped with the cyclization of the PAN block. Indeed, as confirmed by TGA, the PBA entirely decomposed at temperatures below 400 °C (Figure S4). This indicates that temperatures below 300 °C as well as slow heating rates should be maintained during the stabilization step in order to effectively cross-link PAN before the onset of PBA degradation. Nanostructure of BCPs and Mesoporous Carbons. Figure 4 shows the TEM images of CTNCs prepared from
Figure 2. GPC traces for chain extension of (AN)86-Br with BA to yield (AN)86-b-BA54 (Table 2, entry 2).
macroinitiator with higher MW (DP = 184, Table 2, entry 4) partially lost solubility upon addition of BA but slowly dissolved over the course of reaction. This led to a slightly broader MWD (1.29); however, even high MW PAN was successfully chain extended to form a well-defined BCP. Additionally, a “one-pot” synthesis of a PAN-b-PBA BCP via SARA ATRP, without isolating the macroinitiator, was realized (Figure S3). Thermal Analysis of PAN-b-PBA Copolymers. The thermal properties of PAN-containing BCPs illustrate the behavior of precursors during their transformation into nanostructured carbon. PAN-b-PBA BCPs undergo two main thermal processes during carbonization, namely cyclization/ cross-linking of the PAN block and decomposition of the sacrificial PBA block. The cross-linking of PAN is crucial for producing well-defined nanocarbons, since it stabilizes the selfassembled morphology before degradation of the sacrificial PBA block can occur. Too high stabilization temperature can lead to collapse of the structure if a BCP reaches its order− disorder transition temperature before sufficient cross-linking of the PAN phase has occurred. Upon heating in nitrogen, the PAN first experiences cyclization, resulting in a sharp exotherm (peak position: ∼303−310 °C), followed by a relatively broad endotherm (peak position: ∼396 °C) assigned to the decomposition of PBA (Figure 3a). In all cases the exotherms were well separated from the endotherms, suggesting that cyclization and decomposition processes occurred independently in respective blocks during carbonization. This is highly desirable since PBA plays an auxiliary role as sacrificial block and should minimally interfere with cross-linking of PAN domains. However, the DSC thermograms recorded under air revealed two broad, partially overlapping exotherms. The first one with a
Figure 4. TEM images of mesoporous nanocarbons prepared by pyrolysis of (a) (AN)70-b-(BA)40, (b) (AN)86-b-(BA)54, (c) (AN)122-b(BA)79, and (d) (AN)184-b-(BA)124.
Figure 3. DSC thermograms of PAN-b-PBA BCPs with different MWs recorded under (a) nitrogen and (b) air. D
DOI: 10.1021/acs.macromol.6b02678 Macromolecules XXXX, XXX, XXX−XXX
Article
Macromolecules Table 3. Specific Surface Areas and Pore Size Distribution of Mesoporous Carbons Prepared from PAN-b-PBA Block Copolymers specific surface area (m2/g)
a
sample
PAN (wt %)
microporesa (
