Full Terms & Conditions of access and use can be found athttps://www.tandfonline.com/action/journalInformation?journalCode=ynan20

Nanocomposites

ISSN: 2055-0324 (Print) 2055-0332 (Online) Journal homepage: https://www.tandfonline.com/loi/ynan20

Mechanical properties of polymer graftednanoparticle composites

Marissa Giovino, Julia Pribyl, Brian Benicewicz, Ronald Bucinell & LindaSchadler

To cite this article: Marissa Giovino, Julia Pribyl, Brian Benicewicz, Ronald Bucinell &Linda Schadler (2018) Mechanical properties of polymer grafted nanoparticle composites,Nanocomposites, 4:4, 244-252, DOI: 10.1080/20550324.2018.1560988

To link to this article: https://doi.org/10.1080/20550324.2018.1560988

© 2019 The Author(s). Published by InformaUK Limited, trading as Taylor & FrancisGroup.

View supplementary material

Published online: 19 Jan 2019.

Submit your article to this journal

Article views: 156

View Crossmark data

Citing articles: 1 View citing articles

RESEARCH ARTICLE

Mechanical properties of polymer grafted nanoparticle composites

Marissa Giovinoa, Julia Pribylb, Brian Benicewiczb, Ronald Bucinellc and Linda Schadlerd

aMaterials Engineering, Rensselaer Polytechnic Institute, Troy, NY, USA; bChemistry and Biochemistry, University of SouthCarolina, Columbia, SC, USA; cMechanical Engineering, Union College, Schenectady, NY, USA; dMechanical Engineering, Universityof Vermont, Burlington, VT, USA

ABSTRACTPolymer nanocomposites have improved mechanical, optical and thermal properties com-pared to traditional thermoplastics. To study the impact of grafted brush–matrix interactionson the mechanical properties of nanoparticle filled polymers, uniaxial tensile testing withdigital image correlation (DIC) was done on polystyrene (PS) grafted SiO2 in low molecularweight (MW) (N � P), N¼degree of polymerization of grafted brush, P ¼ degree of polymer-ization of matrix, and high MW (2N¼ P) matrices at different loadings. The low matrix MWcomposites had higher strength than high MW composites at high loading but lowerstrength than the pure matrix. The high matrix MW composites had higher toughness atlow loadings. SEM images of fracture surfaces revealed particle debonding and plastic voidgrowth in the high toughness samples. The results show that the P/N ratio of polymergrafted nanoparticle composites is important for controlling the mechanical properties ofpolymer nanocomposites.

ARTICLE HISTORYReceived 2 December 2018Accepted 17 December 2018

KEYWORDSNanocomposite; toughen-ing; mechanical properties;particle debonding; polymermatrix; plastic void growth;grafted brush; RAFTpolymerization

1. Introduction

Polymer nanocomposites often exhibit improvedmechanical properties,1–4 optical properties,5,6 ther-mal stability7–9 and electrical properties10–12 overtheir polymer counterparts. The filler size, distribu-tion and dispersion can change each of those prop-erties significantly.13–16 It has been shown that forlow filler loadings, composites with well dispersedfillers are most likely to exhibit enhanced proper-ties.17 For example, in well dispersed 40 nm aluminafilled PMMA, increases of an order of magnitude in

fracture strain were observed.18 While for clusteredcomposites with 40 nm Al2O3 and 90 nm Fe3O4 inPMMA and PS,19 the modulus of the compositeswas slightly lower than pure polymer in allcomposites measured, which is opposite to what ispredicted by filler reinforcement theories.20,21

Short molecule modification can lead toimproved dispersion, and there are many examplesof well dispersed nanoparticles increasing themodulus.20,21 It is also typical that there is anoptimal loading to achieve the highest toughness or

�CONTACT Marissa Giovino marissa.giovino@gmail.com Materials Engineering, Rensselaer Polytechnic Institute, Troy, NY, USASupplemental data for this article can be accessed here.

� 2019 The Author(s). Published by Informa UK Limited, trading as Taylor & Francis Group.This is an Open Access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/4.0/), which permitsunrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.

NANOCOMPOSITES2018, VOL. 4, NO. 4, 244–252https://doi.org/10.1080/20550324.2018.1560988

impact energy. For example, in 16 nm diameterchlorosilane modified particles in polystyrene,22 themodulus increased with SiO2 loading. The highesttoughness, impact energy and fracture strain werefound at 0.75 wt% SiO2 and SEM images revealedsome evidence of particle debonding.

To further improve dispersion, grafted polymerchains have been used. A bimodal population ofbrushes (one group is densely grafted short brushand the other is sparsely grafted long brush) havebeen found to be optimal for controlling disper-sion.17,23,24 At low loadings, bimodal samples havethe largest modulus in both compression and ten-sion compared to monomodal brush modified par-ticles. In fact the bimodal samples have stiffnesseven higher than predicted by the Guth–Gold20 orHalpin–Tsai21 predictions.

The interactions between grafted brush andmatrix polymer are also important in determiningproperties. In samples without any matrix added,the toughest samples had the highest brush molecu-lar weight (and the most entanglements).25 Notsurprisingly, in samples with matrix added, thelargest strain to failure was also for samples withthe greatest degree of entanglement.26,27 Theseexperimental observations are supported by compu-tational work. Self-consistent field theory (SCFT)has shown that when the matrix and particles arein the wetting regime, the matrix chains penetratethe grafted corona. To quantify wettability, theinterfacial tension of matrix and grafted brush, cb/h,was used.28 It was found that cb/h > 0 for all P(degree of polymerization of matrix) > N (degreeof polymerization of grafted chain) ratios, while forP < N, cb/h was too small to cause film dewetting.To further understand that interaction, the penetra-tion width, w, of the grafted chain has also beendetermined computationally. Matrix penetrationinto the brush is largest for brushes with a graftdensity in the semi-dilute regime. And as P/Nincreases, w decreases.29 Further, if the brush-matrix interactions are repulsive,30 interactiondecreases and leads to a lower yield strength thanpure matrix. For attractive filler-matrix interactions,computation predicts that the yield strength will bethe same as the matrix. For grafted chains, largerthan 5 Me the craze failure mechanism is predictedto be bond breaking, whereas for matrix chainswhich have lower tension (since both ends aren’tfixed) the transition to bond breaking occurs at20 Me. This experimental and computational worksuggest that the fracture behavior of polymer nano-composites depends on the brush–matrix interac-tions which are controlled by the molecular weightand graft density of the brush.

The challenge experimentally has been that it isdifficult to separately alter the matrix-brush inter-action for a fixed dispersion. Bimodal graftedbrushes allow independent modification of the inter-action while maintaining good dispersion. Thus, wehave developed a materials system of bimodal poly-styrene grafted SiO2 nanoparticles dispersed in poly-styrene. The two sets of composites have P/N ratiosof 1 and 2. We find that at specific P/N ratios andloadings, the mechanical properties are alteredresulting in significantly increased strain to failure.

2. Experimental

2.1. SiO2 8k 110k synthesis

Bimodal brush particles were synthesized using asequential reversible addition-fragmentation chaintransfer (RAFT) polymerization similar to a processdescribed previously.17 To a solution of colloidal sil-ica particles (Nissan Chemicals Inc., 30 wt% disper-sion in MIBK, SiO2 density = 2.2 g/mL) diluted withTHF, added 3-aminopropyldimethylethoxysilane anda trace amount of octyldimethylmethoxysilane tomaintain dispersibility through this synthetic step.This mixture was heated at 65 �C for 4 h under aninert (N2) atmosphere. The surface-anchored aminegroups were then reacted with 2-mercaptothiazoline-activated 4-cyanopentanoic acid dithiobenzoate(CPDB). The graft density of these covalently boundchain transfer agents was determined by comparing aUV–Vis spectrum of a grafted particle sample dis-persed in THF to a calibration curve constructedfrom known amounts of free CPDB in solution. Thesurface polymerization of styrene monomer was per-formed at 65 �C. The polystyrene-grafted particleswere precipitated in hexane and recovered by centri-fugation. The chains from a small sample of polystyr-ene-grafted particles were cleaved using hydrofluoricacid (HF), and the chain length and dispersity wereanalyzed by gel-permeation chromatography (GPC).The remainder of the sample was redispersed inTHF. A large excess of azobis(isobutyronitrile)(AIBN) was used to cleave the RAFT agent to preventfurther chain growth.

The particles, which have been grafted with theshort brush of polystyrene, were refunctionalizedwith CPDB as described above. The graft density ofthe second population of chain transfer agent wasdetermined by normalizing the % silica in the sam-ple based on TGA weight loss due to grafted poly-styrene, and this corrected mass was correlated withthe UV–Vis spectrum of the sample to determinethe density of RAFT agents on the silica surface.The polymerization of this second population ofpolystyrene chains was conducted as described

NANOCOMPOSITES 245

above. After the polystyrene-grafted particles wereprecipitated in hexane, a small sample was takenand the chains cleaved with HF. The molecularweight and dispersity of both populations of cleavedchains was analyzed by GPC. The bimodal polymergrafted particles were redispersed in THF, and thesecond population of RAFT agent was cleaved usinga large excess of AIBN before further use.

The bimodal polystyrene grafted SiO2 is referredto as SiO2 8k 110k in this study. Two populationsof polymer brushes were grafted to the surface: ashort polystyrene brush with MW = 8 kg/mol and agraft density = 0.25 chains/nm2, and long polystyr-ene brush with MW = 110 kg/mol and a graft dens-ity = 0.07 chains/nm2.

2.2. SiO2–PS composites

SiO2 8k 110k in THF was solution mixed withhomopolymer matrix (MW = 100 kg/mol or MW =200 kg/mol, from Polymer Source). The sampleswere mixed with a probe sonicator (Sonics andMaterials Vibracell VCX 750 W unit) for 1 min at40% amplitude using the pulsed setting of 2 s onand 0.5 s off. The samples were solution cast intoaluminum boats at 90 �C. Composites wereannealed for 2 days at 120 �C in a vacuum oven.Samples were hot pressed at 190 �C andslow cooled.

2.3. Transmission electron microscopy (TEM)

Composite samples were embedded in epoxy.Samples were cut into 60 nm sections using anRMC PowerTome XL microtome. The sections werefloated onto water and transferred to copper grids.The sections were imaged using a JEOL 2011 TEMat an accelerating voltage of 200 kV. The imageswere converted to binary format using programImageJ. For each composite 30 images were takenand binarized. To measure cluster size an algorithmdeveloped by the Brinson and Chen groups atNorthwestern University was used.31–33

2.4. Differential scanning calorimetry (DSC)

Composite samples were weighed into aluminumpans for determination of glass transition tempera-ture (Tg). The sample and an empty reference panwere loaded into a DSC-Q100. The sample chamberwas filled with N2 (g). Composites were heated at10 �C/min to 160 �C and cooled to 50 �C at a rateof 10 �C/min. Three heating and cooling cycles wererun. The glass transition temperature (Tg) was takenas the inflection point of the step.

2.5. Thermal gravimetric analysis (TGA)

For SiO2–PS composites, a TGA-Q50 was used todetermine the filler loading. The sample was loadedinto an alumina crucible. The sample chamber wasfilled with N2 (g). The sample was heated from 25to 800 �C at a rate of 20 �C/min.

2.6. Tensile testing with digital imagecorrelation (DIC)

Eight different samples were made: PS matrix 100k,PS matrix 200k, 0.7 vol.% SiO2 in 100k PS, 1.3vol.% SiO2 in 100k PS, 5.5 vol.% SiO2 in 100k PS,0.5 vol.% SiO2 in 200k PS, 1.3 vol.% SiO2 in 200kPS, and 5.5 vol.% SiO2 in 200k PS. For each sample5 specimens were made. Samples were hot pressedat 190 �C using a Carver hot press into dog bonesamples with dimensions specified by ASTM D638type V.

The PS and SiO2–PS composites were preparedfor 3D DIC measurements by applying a randomspeckle pattern.34 The pattern was applied in twosteps: first a uniform white base coat (white enamelpaint (Model Master FS37875) with some air brushthinner) was spray painted. Next a random blackspeckle pattern (Model Master FS37038) wasapplied. The samples and setup are shown inFigure 1. Samples were tested in uniaxial tensionusing an electro-mechanical load frame (load cellcapacity 1 kN) at room temperature. A preload of 5

Figure 1. Tensile testing with DIC. (a) Samples after test with speckle pattern and (b) tensile testing load frame, gripsand cameras.

246 M. GIOVINO ET AL.

N–7 N was applied to each sample. The sampleswere strained at a strain rate of 1 mm/min. The cali-brated field of view was 15 mm width � 10 mmheight � 5 mm depth of field. Images were capturedusing 2 CCD cameras (2448 � 2050 pixel resolution)with Schneider UNIFOC 2.8/50 lenses. The load datawas taken from the load cell. The ARAMIS softwareby GOM was used to correlate the images and calcu-late strains. Virtual extensometers were placed onimages to compute strain. A virtual extensometerapproximately 7 mm in length with a precision of0.001 mm was used to calculate axial strain.

2.7. Scanning electron microscopy (SEM)Fractography

The SiO2–PS composites and PS matrices aftertensile testing were preserved. The fracturesurfaces were sputter coated with 60% Au/40% Pdusing a Hummer V sputter coater. The metal layerwas 0.5 nm thick. Samples were then imaged withFEI Versa 3D Dual Beam SEM at acceleratingvoltage of 3 kV.

3. Results and discussion

The materials system details are shown in Table 1.The polymer grafted nanoparticles in 100k PS arepredicted to have matrix penetration into thegrafted brush since N�P, however the grafted chains

in 200k PS are expected to have more collapsedchains or a smaller amount of matrix penetrationsince N < P.28,29

To evaluate the dispersion state of SiO2–PS com-posites, TEM was used. The images were binarizedand 30 images were taken per composite. An algo-rithm was applied to the binarized images to extractaverage cluster radius (Rc).

31–33 Representative TEMimages for the SiO2–PS composites are shown inFigure 2. Each image clearly shows that the particlesare singly dispersed and well distributed throughoutthe micrograph. To quantify the dispersion, Rc foreach sample was calculated. For each compositesample, Rc=7 nm as expected for well dispersedpolymer nanocomposites with 14 nm diameter par-ticles. Thus, any differences in properties cannot beattributed to the particle distribution.

The thermal state of polymer nanocomposites isalso important when considering the mechanicalproperties. Specifically, for amorphous polymers theTg is very important. The Tg values are shown inTable 2 below. The 100k PS composites and matrixTg values are within error of each other. The ther-mal state does not change with particle loading for100k samples. For the 200k composites, the largestvariation is less than 2 degrees. Thus, any significantdifference in mechanical properties is not due tochanges in Tg.

Figure 3a shows the modulus data vs filler vol-ume fraction data (the core SiO2 volume fraction).

Table 1. Materials system for bulk mechanical testing of PS grafted SiO2 in homopolymer PS.sample ID Matrix MW (kg/mol) SiO2 loading Surface modification

100k PS 100 – –100k_0.7 100 0.7 vol% PS: 8 kg/mol, 0.25 chains/nm2 PS: 110 kg/mol, 0.07 chains/nm2

100k_1.3 100 1.3 vol% PS: 8 kg/mol, 0.25 chains/nm2 PS: 110 kg/mol, 0.07 chains/nm2

100k_5.5 100 5.5 vol% PS: 8 kg/mol, 0.25 chains/nm2 PS: 110 kg/mol, 0.07 chains/nm2

200k PS 200 – –200k_0.5 200 0.5 vol% PS: 8 kg/mol, 0.25 chains/nm2 PS: 110 kg/mol, 0.07 chains/nm2

200k_1.3 200 1.3 vol% PS: 8 kg/mol, 0.25 chains/nm2 PS: 110 kg/mol, 0.07 chains/nm2

200k_5.5 200 5.5 vol% PS: 8 kg/mol, 0.25 chains/nm2 PS: 110 kg/mol, 0.07 chains/nm2

Figure 2. TEM images of SiO2-PS composites, all scale bars 200 nm. (a) Image of 100k_1.3, (b) Image of 100k_0.7, (c) Imageof 200k_1.3, and (d) Image of 200k_0.5.

NANOCOMPOSITES 247

The prediction from filler reinforcement theory byGuth–Gold is shown for comparison.20 The 100kand 200k composites have very similar values formodulus at constant loading and follow theGuth–Gold prediction closely. There are a fewdeviations from Guth–Gold at very low andhigh loadings.

The strength data vs core SiO2 volume fraction isshown in Figure 3b. The strength of the 100k MWmatrix composites does not change significantlywith loading, while the 200k matrix compositesshow a decrease in strength at higher loading. The100k_0.7 composite failed mostly outside the gaugesection so the data point is from only 2 specimens.Clearly in these systems, the impact of the matrix–brush interaction is not large enough to change themaximum load significantly at low loadings. Athigher loadings, when the brushes have the potentialto overlap, the strength begins to decrease. A plot ofthe interparticle distance is shown in Figure 4. Hereinterparticle distance, dc–c, includes the core SiO2

particle and the grafted brush. The grafted brushheight was calculated as radius of gyration of thelong brush (h = 9.1 nm). This has been shown to bea good approximation of brush height according toSAXS and SANS measurements.35 It is clear fromFigure 4 that at a high particle loading the interpar-ticle distance is very small. This small distance islikely the cause of the strength decrease.

The failure strain data as a function of SiO2

volume fraction is shown in Figure 3c. There is one

surprising data point. For the 200k_0.5 compositethere is a very large strain to failure. This is thesame loading where the decrease in modulus wasobserved. 100k composites at the same loadingtended to fail in the grips, but the few samplestested did not show the increase in strain to failure.The large failure strain and high strength suggestthat the 200k_0.5 composite is tougher than all theother samples. Strain energy density, the area underthe stress–strain curve, was used as a measure oftoughness. The 200k_0.5 composites had twice thestrain energy density compared to the matrix. Onepossible reason for this increased strain energydensity is residual solvent. Residual solvent can actas a plasticizer and toughener.36 All samples wereprocessed in the same way so it is highly unlikelythat the 200k_0.5 composite has residual solvent.Furthermore, if there was residual solvent a decreasein Tg would be seen. In Table 2, the Tg of 200k_0.5is similar to all other samples. To better understandthis large strain energy density, the fracture surfacesof tensile testing samples were imaged with SEM.

The fracture surfaces of the 200k SiO2–PS com-posites were imaged with SEM and are shown inFigure 5. Representative images for low and highmagnification are shown for each sample. The lowmagnification images for each 200k sample are verysimilar. The high magnification images have signifi-cant differences. In Figure 5b the matrix high mag-nification image is shown for reference, the fracturesurface is very smooth with minimal features. Thehigh magnification images for the 200k_1.3 and200k_5.5 composites are very similar to each otherand slightly different than the matrix image (Figure5f and h). The SEM images for 200k_1.3 and200k_5.5 have some small pores which could befrom particle debonding. The average pore size forFigure 5f is 46 nm diameter and for Figure 5h it is41 nm diameter. These voids are roughly 3 times

Figure 3. Bulk mechanical properties data for polymer nanocomposites. Data points with asterisk only had 2 samples break ingauge length. (a) Modulus vs. particle loading for polymer nanocomposites, (b) Failure strength vs. particle loading forpolymer composites, and (c) failure strain vs. particle loading for polymer nanocomposites.

Table 2. The Tg for polystyrene matrices and SiO2-PS com-posites. The composites are labeled with vol.% of SiO2 only.Sample name Tg (�C) Sample name Tg (�C)100k PS 105.8 200k PS 107.3100k_0.7 106 200k_0.5 106.9100k_1.3 106.1 200k_1.3 107.8100k_5.5 105.4 200k_5.5 105.4

248 M. GIOVINO ET AL.

larger than the SiO2 diameter. The SEM images for200k_0.5 have very large pores (Figure 5d), muchlarger than the particle size, 2R = 14 nm, likely therewas particle debonding followed by plastic voidgrowth. If we look more closely at the fracture sur-face of composite 200k_0.5 we note that there aretwo regions (Figure 6). There are more voids andlarger voids in region 1 than region 2. Region 1appears to have a rough fracture surface with a large

amount of deformation. Region 1 is likely the slowcrack growth region, providing more time forvoid growth, and region 2 is the fast crackgrowth region.

To further understand the large fracture strainobserved for composite 200k_0.5 several tougheningmechanisms were considered. Common tougheningmechanisms for particle filled composites are: crackpinning, crack deflection, shear banding and plasticvoid growth.37–40 In crack pinning the particles actas obstacles that the crack will bow around. Thecrack bowing increases the crack length, which,increases toughness. Our obstacles are 14 nm SiO2

nanoparticles. In most cases of crack pinning theobstacles were larger than the crack tip opening dis-placement:41–44

dtc ¼ K21c

Ery1�t2ð Þ (1)

Figure 5. SEM images of fractured surfaces (a–b) Images of 200k PS, (c–d) Images of 200k_0.5, (e–f) Images of 200k_1.3, and(g–h) Images of 200k_5.5.

Figure 4. Interparticle distance vs particle loading. The par-ticle is defined as core inorganic SiO2 and grafted PS brush.

NANOCOMPOSITES 249

where dtc is crack opening displacement, K1c isfracture toughness, E is Young’s modulus, ry is yieldstress and � is Poisson’s ratio. Using the values inTable 3, dtc =6.1 um, this is more than 2 orders ofmagnitude larger than our particle diameter. It ishighly unlikely that the particles are crack pinningobstacles. Next, we consider crack deflection as atoughening mechanism. In crack deflection, thecrack is tilted and twisted around the particles,which, causes an increase in the fracture surfacearea and crack growth occurs in mixed mode frac-ture (mode I and mode II).39,45 The particles arelikely too small to cause this effect. According toEvans et al., the particles need to be larger than theplastic zone radius in order for this mechanism toapply.45,46 To calculate plastic zone radius ry thefollowing equation was used:

ry ¼ 16p

K1c

ry

� �2

(2)

where ry=24.5 mm. The particle diameter is a fac-tor of 1000 smaller than this indicating that crackdeflection is not likely for these fillers. Shear band-ing is another common toughening mechanism forparticle filled composites. There was no obviousnecking of the sample or evidence of 45� bands inthe DIC data or fracture surface. Suggesting there isno significant shear banding in these samples.

The final mechanism to consider is plastic voidgrowth. This mechanism starts with particledebonding from the matrix and is followed bygrowth of voids. The void growth is the dominantenergy dissipation mechanism. To model this thefollowing equations were used:37

DGcomp ¼ DGm þ DGv (3)

DGv ¼ 1�l2m3

� �Vv�Vfð ÞrycryK2

vm (4)

where DGcomp is the strain energy release rate ofcomposite, DGm is the strain energy release rate ofthe matrix, DGv is the energy dissipated from plasticvoid growth, lm is von Mises pressure sensitivity, Vv

is void volume fraction, Vf is filler volume fraction,ryc is compressive yield stress, and Kvm is the vonMises stress concentration factor. The propertiesused in this calculation are shown in Table 3.38,47,48

Instead of strain energy release rate, we used thestrain energy density, area under stress– strain curve,for composite (Uc) and matrix (Um), Uc/Um=2.6. Weassumed that this strain energy density ratio is equalto the strain energy release rate ratio: DGcomp/DGm

ratio. A maximum DGv was calculated by assumingall the particles debonded and knowing the averagevoid diameter is 50 nm (Figure 6c). The calculationfound DGv/DGm > 1.6 indicating 100% particledebonding is an overestimate. Partial particledebonding followed by plastic void growth is the bestexplanation for the high strain energy density seen inthe 200k_0.5 composite.

4. Conclusions

The effect of grafted chain conformation on bulkmechanical properties of SiO2–PS composites was

Figure 6. SEM images of fracture surface for composite 200k_0.5. (a) Low magnification image of fracture surface with region1 and 2 labeled, (b) high magnification image of region 2, and (c) high magnification image of region 1.

Table 3. Properties used in evaluation of toughen-ing mechanisms.Property Reference Property References

E 3.14 GPa – lm 0.25 47

� 0.33 – Kvm 2.22 37

ry 41.9MPa – ryc 70MPa 48

K1c 0.9MPaffiffiffiffim

p 47 ry 24.5 um –

The – in reference column indicates it was measured experimentally.

250 M. GIOVINO ET AL.

tested using tensile testing with DIC. Two compositesystems were used: one with N � P and one with N> P. For N � P the grafted chains are expected tobe swollen and increase the composite strength,whereas when N < P the grafted chains areexpected to be collapsed and facilitate particle pull-out. The composites with N � P had roughly con-stant strength with particle loading and even higherstrength than 200k_5.5. The N < P composite hadhigh fracture strain and strength at low loadings,which suggests higher toughness. Toughness esti-mated with strain energy density for composite200k_0.5 was twice as large as 200k matrix. SEMimages of fracture surfaces suggested that the largestrain energy density was due to particle debondingfollowed by plastic void growth. The high magnifi-cation images showed two zones: 1 a slow crackgrowth zone containing a lot of deformation andlarge voids and 2 a fast crack growth zone with lessdeformation and smaller voids. The bulk mechanicalproperties of polymer nanocomposites were system-atically studied, it was found that composite P/Nratio can be tuned to increase strength or strainenergy density.

Acknowledgments

MG acknowledges funding from the NSF under coopera-tive agreement EEC-0812056, New York State underNYSTAR contract C090145 and the materials engineeringdepartment at Rensselaer Polytechnic Institute.

Disclosure statement

No potential conflict of interest was reported bythe authors.

Notes on contributors

Marissa Giovino is a materials engineer at the Air ForceResearch Lab in WPAFB, Ohio. She recently completedher PhD at Rensselaer Polytechnic Institute. Her thesiswas focused on the mechanical and viscoelastic propertiesof polymer nanocomposites.

Julia Pribyl received her PhD from the University ofSouth Carolina where she focused on synthesis of poly-mer-grafted nanomaterials and foams. She is now a post-doctoral associate at the University of Florida.

Brian C. Benicewicz joined the University of SouthCarolina, where he holds the SmartState Chair in PolymerNanocomposites in the Department of Chemistry andBiochemistry. His research focuses on the development ofhigh temperature membranes for fuel cells, reversible add-ition-fragmentation chain transfer (RAFT) polymeriza-tion, and the preparation of multifunctional nanoparticlesand polymer nanocomposites.

Ronald B. Bucinell is an associate Professor ofMechanical Engineering at Union College, a Fellow of theAmerican Society of Mechanical Engineers, and a licensed

Professional Engineer in the state of New York. His areasof interest are experimental mechanics and compos-ite materials.

Linda Schadler is the Dean of Engineering andMathematical Sciences at the University of Vermont andprofessor in the Mechanical Engineering Department. Herresearch focuses on the dielectric, mechanical, and opticalproperties of polymer nanocomposites.

References

1. Shah D, Maiti P, Jiang DD, et al. Effect of nanopar-ticle mobility on toughness of polymer nanocompo-sites. Adv Mater. 2005;17:525–528.

2. Zhang H, Zhang Z, Friedrich K, et al. Propertyimprovements of in situ epoxy nanocompositeswith reduced interparticle distance at high nanosil-ica content. Acta Mater. 2006;54:1833–1842.

3. Wetzel B, Haupert F, Zhang MQ. Epoxy nanocom-posites with high mechanical and tribological per-formance. Compos Sci Technol. 2003;63:2055–2067.

4. Zhao X, Zhang Q, Chen D, et al. Enhanced mech-anical properties of graphene-based polyvinyl alco-hol composites. Macromolecules. 2010;43:2357–2363.

5. Dang A, Ojha S, Hui CM, et al. High-transparencypolymer nanocomposites enabled by polymer-graftmodification of particle fillers. Langmuir. 2014;30:14434–14442.

6. Kim J, Yang H, Green PF. Tailoring the refractiveindices of thin film polymer metallic nanoparticlenanocomposites. Langmuir. 2012;28:9735–9741.

7. Zhu J, Uhl FM, Morgan AB, et al. Studies on themechanism by which the formation of nanocompo-sites enhances thermal stability. Chem Mater. 2001;13:4649–4654.

8. Wen X, Wang Y, Gong J, et al. Thermal andflammability properties of polypropylene/carbonblack nanocomposites. Polym Degrad Stab. 2012;97:793–801.

9. Gilman J. Flammability and thermal stability studiesof polymer layered-silicate (clay) nanocomposites.Appl Clay Sci. 1999;15:31–49.

10. Wang J, Dai J, Yarlagadda T. Carbon nanotube -conducting-polymer composite nanowires.Langmuir. 2005;21:9–12.

11. Geng Y, Liu MY, Li J, et al. Effects of surfactanttreatment on mechanical and electrical properties ofCNT/epoxy nanocomposites. Compos Part A: ApplSci Manuf. 2008;39:1876–1883.

12. G�omez H, Ram MK, Alvi F, et al. Graphene-con-ducting polymer nanocomposite as novel electrodefor supercapacitors. J Power Sources. 2011;196:4102–4108.

13. Chevigny C, Dalmas F, Di Cola E, et al. Polymer-grafted-nanoparticles nanocomposites: dispersion,grafted chain conformation, and rheological behav-ior. Macromolecules. 2011;44:122–133.

14. Akcora P, Kumar SK, Moll J, et al. Gel-like mech-anical reinforcement in polymer nanocompositemelts. Macromolecules. 2010;43:1003–1010.

15. Moll JF, Akcora P, Rungta A, et al. Mechanicalreinforcement in polymer melts filled with polymergrafted nanoparticles. Macromolecules. 2011;44:7473–7477.

NANOCOMPOSITES 251

16. Akcora P, Liu H, Kumar SK, et al. Anisotropic self-assembly of spherical polymer-grafted nanoparticles.Nat Mater. 2009;8:354–359.

17. Natarajan B, Neely T, Rungta A, et al.Thermomechanical properties of bimodal brushmodified nanoparticle composites. Macromolecules.2013;46:4909–4918.

18. Ash BJ, Rogers DF, Wiegand CJ, et al. Mechanicalproperties of Al2O3/polymethylmethacrylate nano-composites. Polym Compos. 2002;23:1014–1025.

19. Ciprai D, Jacob K, Tannenbaum R. Characterizationof polymer nanocomposite interphase and itsimpact on mechanical properties. Macromolecules.2006;39:6565–6573.

20. Guth E. Theory of filler reinforcement. J Appl Phys.1945;16:20–25.

21. Halpin JC, Kardos JL. The Halpin-Tsai equations: areview. Polym Eng Sci. 1976;16:344–352.

22. Vaziri H, Abadyan M, Nouri M, et al. Investigationof the fracture mechanism and mechanical proper-ties of polystyrene/silica nanocomposite in varioussilica contents. J Mater Sci. 2011;46:5628–5638.

23. Li Y, Krentz TM, Wang L, et al. Ligand engineeringof polymer nanocomposites: from the simple to thecomplex. ACS Appl Mater Interfaces. 2014;6:6005–6021.

24. Li Y, Tao P, Viswanath A, et al. Bimodal surfaceligand engineering: the key to tunable nanocompo-sites. Langmuir. 2013;29:1211–1220.

25. Zheng Y, Abbas ZM, Sarkar A, et al. Surface-initi-ated reversible addition-fragmentation chain trans-fer polymerization of chloroprene and mechanicalproperties of matrix-free polychloroprene nanocom-posites. Polymer. 2018;135:193–199.

26. Donald AM, Kramer EJ. Effect of molecular entan-glements on craze microstructure in glassy poly-mers. J Polym Sci Polym Phys Ed. 1982;20:899–909.

27. Kramer EJ, Berger LL. Fundamental processes ofcraze growth and fracture. In: H.-H. Kausch, editor.Crazing in polymer. Berlin, Heidelberg: Springer;1990.

28. Matsen MW, Gardiner JM. Autophobic dewettingof homopolymer on a brush and entropic attractionbetween opposing brushes in a homopolymermatrix. J Chem Phys. 2001;115:2794–2804.

29. Trombly DM, Ganesan V. Curvature effects uponinteractions of polymer-grafted nanoparticles inchemically identical polymer matrices. J ChemPhys. 2010;133:154904.

30. Meng D, Kumar SK, Ge T, et al. Crazing ofnanocomposites with polymer-tethered nanopar-ticles. J Chem Phys. 2016;145:094902

31. Hassinger I, Li X, Zhao H, et al. Toward thedevelopment of a quantitative tool for predictingdispersion of nanocomposites under non-equilib-rium processing conditions. J Mater Sci. 2016;51:4238–4249.

32. Zhang Y, Zhao H, Hassinger I, et al. Microstructurereconstruction and structural equation modeling forcomputational design of nanodielectrics. IntegrMater Manuf Innov. 2015;4: 14.

33. Xu H, Li Y, Brinson C, et al. A descriptor-baseddesign methodology for developing heterogeneousmicrostructural materials system. J Mech Des. 2014;136:051007.

34. Schreier H, Orteu J-J, Sutton MA. Image correlationfor shape, motion and deformation measurements.Boston, MA: Springer US; 2009.

35. Chevigny C, Jestin J, Gigmes D, et al. Wet-to-dryconformational transition of polymer layersgrafted to nanoparticles in nanocomposite.Macromolecules. 2010;43:4833–4837.

36. Eriksson M, Goossens H, Peijs T. Influence of dry-ing procedure on glass transition temperature ofPMMA based nanocomposites. Nanocomposites.2015;1:36–45.

37. Johnsen BB, Kinloch AJ, Mohammed RD, et al.Toughening mechanisms of nanoparticle-modifiedepoxy polymers. Polymer. 2007;48:530–541.

38. Huang Y, Kinloch AJ. Modelling of the tougheningmechanisms in rubber-modified epoxy polymerspart II: a quantitative description of the microstruc-ture-fracture property relationships. J Mater Sci.1992;27:2763–2769.

39. Huang Y, Kinloch AJ. Modelling of the tougheningmechanisms in rubber-modified epoxy polymerspart 1: finite element analysis studies. J Mater Sci.1992;27:2753–2762.

40. Lange FF. The interaction of a crack front with asecond-phase dispersion. Philos Mag. 1970;22:983–992.

41. Green D, Nicholson P, Embury J. Fracture of a brit-tle particulate composite, part 2: theoretical aspects.J Mater Sci. 1979;14:1657–1661.

42. Kinloch AJ, Maxwell DL, Young RJ. The fracture ofhybrid-particulate composites. J Mater Sci. 1985;20:4169–4184.

43. Lange FF, Radford KC. Fracture energy of an epoxycomposite system. J Mater Sci. 1971;6:1197–1203.

44. Kinloch AJ, Young RJ. Fracture behaviour of poly-mers. New York: Applied Science Publishers; 1983.

45. Evans AG, Williams S, Beaumont PWR. On thetoughness of particulate filled polymers. J Mater Sci.1985;20:3668–3674.

46. Evans AG. The strength of brittle materials contain-ing second phase dispersions. Philos Mag. 1972;26:1327–1344.

47. S. Lampman, editor. Characterization and failureanalysis of plastics. Singapore: ASM International;2003.

48. Wu JJ, Buckley CP. Plastic deformation of glassypolystyrene: a unified model of yield and the role ofchain length. J Polym Sci B: Polym Phys. 2004;42:2027–2040.

252 M. GIOVINO ET AL.