http://jba.sagepub.com/Journal of Biomaterials Applications

http://jba.sagepub.com/content/29/2/209The online version of this article can be found at:

 DOI: 10.1177/0885328214521252

2014 29: 209 originally published online 30 January 2014J Biomater ApplRoss W Ormsby, Mircea Modreanu, Christina A Mitchell and Nicholas J Dunne

Carboxyl functionalised MWCNT/polymethyl methacrylate bone cement for orthopaedic applications  

Published by:

http://www.sagepublications.com

can be found at:Journal of Biomaterials ApplicationsAdditional services and information for    

  http://jba.sagepub.com/cgi/alertsEmail Alerts:

 

http://jba.sagepub.com/subscriptionsSubscriptions:  

http://www.sagepub.com/journalsReprints.navReprints:  

http://www.sagepub.com/journalsPermissions.navPermissions:  

http://jba.sagepub.com/content/29/2/209.refs.htmlCitations:  

What is This? 

- Jan 30, 2014OnlineFirst Version of Record  

- Jul 10, 2014Version of Record >>

at UNIVERSITE LAVAL on July 16, 2014jba.sagepub.comDownloaded from at UNIVERSITE LAVAL on July 16, 2014jba.sagepub.comDownloaded from

Article

Carboxyl functionalised MWCNT/polymethyl methacrylate bonecement for orthopaedic applications

Ross W Ormsby1, Mircea Modreanu2, Christina A Mitchell3 and Nicholas J Dunne1

Abstract

The incorporation of carboxyl functionalised multi-walled carbon nanotube (MWCNT-COOH) into a leading propri-

etary grade orthopaedic bone cement (Simplex PTM) at 0.1 wt% has been investigated. Resultant static and fatigue

mechanical properties, in addition to thermal and polymerisation properties, have been determined. Significant improve-

ments (p� 0.001) in bending strength (42%), bending modulus (55%) and fracture toughness (22%) were demonstrated.

Fatigue properties were improved (p� 0.001), with mean number of cycles to failure and fatigue performance index

being increased by 64% and 52%, respectively. Thermal necrosis index values at �44�C and �55�C were significantly

reduced (p� 0.001) (28% and 27%) versus the control. Furthermore, the onset of polymerisation increased by 58%

(p< 0.001), as did the duration of the polymerisation reaction (52%). Peak energy during polymerisation increased by

672% (p< 0.001). Peak area of polymerisation increased by 116% (p< 0.001) indicating that the incorporation of

MWCNT-COOH reduced the rate of polymerisation significantly. A non-significant reduction (8%) in percentage mono-

mer conversion was also recorded. Raman spectroscopy clearly showed that the addition of MWCNT-COOH increased

the ratio between normalised intensities of the G-Band and D-Band (IG/ID), and also increased the theoretical compres-

sive strain (�1.72%) exerted on the MWCNT-COOH by the Simplex PTM cement matrix. Therefore, demonstrating a

level of chemical interactivity between the MWCNT-COOH and the Simplex PTM bone cement exists and consequently

a more effective mechanism for successful transfer of mechanical load. The extent of homogenous dispersion of the

MWCNT-COOH throughout the bone cement was determined using Raman mapping.

Keywords

Polymethyl methacrylate bone cement, multi-walled carbon nanotube, mechanical properties, thermal properties, Raman

spectroscopy

Introduction

Polymethyl methacrylate (PMMA) bone cement is usedas a load-transferring grout between the implant pros-thesis and bone in joint replacement surgery. The poly-merisation reaction of PMMA bone cement is highlyexothermic resulting in temperatures being generated inexcess of 100�C. These elevated temperatures can causesignificant cellular bone necrosis and potentially asepticloosening.1,2 The number of revision total knee replace-ments performed in the US per annum will increase six-fold to 270,000 by 2030 and the number of total hipreplacement revisions will increase approximately two-fold to 97,000, with failure of the cement mantle beingcited as the most prevalent cause of implant failure.1,3

It has been successfully demonstrated that carbonnanotube (CNT) incorporation can improve the

mechanical, thermal and electrical properties of arange of conventional polymer systems.3–9 Andrewsand Weisenberger7 proposed that the mechanical, ther-mal and electrical property improvements for CNT/polymer composites are a function of CNT type,

Journal of Biomaterials Applications

2014, Vol. 29(2) 209–221

! The Author(s) 2014

Reprints and permissions:

sagepub.co.uk/journalsPermissions.nav

DOI: 10.1177/0885328214521252

jba.sagepub.com

1School of Mechanical and Aerospace Engineering, Queen’s University of

Belfast, Belfast, UK2Micro/Nanoelectronics Department, Tyndall National Institute, Lee

Maltings, Cork, Republic of Ireland3School of Medicine, Dentistry and Biomedical Science, Queen’s

University of Belfast, Belfast, UK

Corresponding author:

Nicholas J Dunne, School of Mechanical & Aerospace Engineering,

Queen’s University of Belfast, Ashby Building, Stranmillis Road, BT9 5AH,

Belfast, UK.

Email: n.dunne@qub.ac.uk

at UNIVERSITE LAVAL on July 16, 2014jba.sagepub.comDownloaded from

degree of dispersion, level of loading, CNT alignmentand polymer matrix.7 The mechanical and thermalproperties of PMMA bone cement have previouslybeen modified with multi-walled carbon nanotube(MWCNT) powder while maintaining biocompatibilityby Ormsby et al.10,11 Reductions in the exothermic tem-perature during polymerisation and thermal necrosisindex (TNI) values for MWCNT-PMMA cement at aweight loading level of 0.1wt% were reported.MWCNT dispersion within the methyl methacrylate(MMA) liquid monomer using ultrasonic disintegrationproved the most efficacious method. Ormsby et al.12

also demonstrated that chemical functionality andweight loading level of the MWCNT significantly influ-enced the fatigue mechanical and thermal properties ofthe same PMMA cement formulation. Ormsby et al.13

showed that MWCNT addition significantly altered thepolymerisation reaction and cure kinetics for the result-ant PMMA bone cement model. The extent of thiseffect was largely governed by MWCNT loading level,chemical functionality and the achievement of homo-genous dispersion within the PMMA microstructure.The rate of the polymerisation reaction was signifi-cantly altered such that the setting time of PMMAcement was extended. This reduced rate of reactionwas attributed to the MWCNT having an active rolein the free radical polymerisation process. The onset ofpolymerisation and critical gelation time varied as afunction of MWCNT type and loading level. It wasalso postulated that the MWCNT network within thePMMA microstructure demonstrated both a chemicaland physical interaction during the polymerisationreaction. These interactions assisted in the dissipationof heat energy generated during polymerisation, in add-ition to prolonging the free radical reaction. Ormsbyet al.13 suggested that the higher level of chemical inter-action between the carboxyl functionalised multi-walled carbon nanotube (MWCNT-COOH) andPMMA during the polymerisation reaction was theprincipal driver for reducing the extent of the exother-mic reaction as the incorporation of unfunctionalisedMWCNT did not have any appreciable influence. Asimilar finding was reported by Goncalves et al.,14

when investigating CNT as reinforcing agent in aPMMA/hydroxyapatite composite bone cement.Goncalves et al. suggested that the CNT could act asa radical scavenger during polymerisation.

It should be noted that the PMMA bone cementused in previous studies was not a proprietary system,but a PMMA-based bone cement analogous to a com-mercially available cement in terms of chemical formu-lation.10–13 Notwithstanding this fact, this PMMAbone cement is not approved for clinical applicationas a grouting material during joint replacement surgery.Therefore, it is essential to characterise and understand

the effect of incorporating MWCNT-COOH into a pro-prietary PMMA bone cement that is routinely used inorthopaedic applications, e.g. a total hip replacementand total knee replacement.

In this present study, we report the influence ofincorporating 0.1wt% MWCNT-COOH into a com-mercially available PMMA bone cement in terms ofthe effect on thermal properties and mechanical per-formance under static and fatigue loading conditions.We have also examined the effect of MWCNT additionon the extent of the polymerisation reaction and mono-mer conversion of the PMMA bone cement using dif-ferential scanning calorimetry (DSC). MWCNTdispersion and the extent of chemical interaction withthe PMMA bone cement matrix was also investigatedusing Raman spectroscopy.

Materials and methods

Materials and preparation of bone cement

Simplex PTM bone cement (Stryker HowmedicaOsteonics, Republic of Ireland) and MWCNT-COOH(4wt% COOH concentration) (Nanocyl S.A., Belgium)were used in this study. Carboxyl functionalisedMWCNT was chosen for this study as it has previouslybeen shown to provide optimal mechanical and thermalproperties for non-commercial PMMA bone cementwhen added at a loading level of 0.1wt%.10–13

MWCNT-COOH (0.1wt%) were incorporated intothe MMA liquid monomer by dispersion, using anultrasonic disintegrator (MSE Ltd. UK) at an ampli-tude of 10� 1 mA for 30� 1 s (X3).10 Subsequently, theMWCNT-MMA monomer suspension and PMMA/styrene polymer powder were mixed together underambient conditions (22� 1�C and at a relative humidityof not less than 40%), using a commercially availablevacuum mixing system (Summit Medical Ltd., UK) inaccordance with the manufacturer’s instructions.Control specimens containing no MWCNT-COOHwere also prepared.

Mechanical characterisation

Static properties. Compressive and bending propertieswere determined in accordance with ISO 5833:2002.15

Fracture toughness was determined using the Chevron-Notch Short Rod (CNSR) technique.11 All mechanicaltesting was conducted using a Universal materials test-ing machine (Lloyd’s Instrument Ltd., UK).10 For eachstatic mechanical property determined, a total of 18bone cement specimens (i.e. six specimens from threeseparate mixes) were tested. Specimens were pre-conditioned under ambient laboratory conditions(22� 1�C) for 24� 0.5 h prior to testing.

210 Journal of Biomaterials Applications 29(2)

at UNIVERSITE LAVAL on July 16, 2014jba.sagepub.comDownloaded from

Fatigue properties. Specimens were prepared by injectingpolymerising bone cement into appropriate polytetra-fluoroethylene (PTFE) moulds and allowed to cure fora minimum of 24� 0.5 h. The dimensions for each fati-gue specimen were 75� 0.5mm in length, 5� 0.2mm inwidth and 3.5� 0.2mm in thickness, with a gaugelength of 25� 0.5mm. Each test specimen was mea-sured and stored for at least one week at 37� 1�C indry conditions prior to testing.16 The fatigue tests wereperformed using a purpose-built, pneumatically con-trolled testing machine (Zwick-Roell, UK) and cycledcontinuously in load control until failure. The cyc-lic stress employed was sinusoidal at a frequency of2Hz. Twenty six samples were tested in tension –tension with a lower stress of 0.3MPa and an upperstress of 22.0MPa. All the testing was conducted inair at 22� 1�C. Each specimen was tested until failureand the maximum number of cycles before failurerecorded.

The fatigue test results were analysed using the fol-lowing methods:

a. Probability of fracture method,17,18

b. Three parameter Weibull method,17–19

c. Probability of survival method,20

d. Fatigue performance index, I, approach.21

Survival analysis using the Weibull relationship hasbeen used in many previous studies to analyse fatiguetest data of bone cement specimens.16,22–26

Thermal characterisation

Setting properties

Maximum temperature (Tmax) and setting time (tset) forsix specimens of each bone cement mix was measured inaccordance with ISO5833:2002.15 Using Simpson’sRule, the area under each temperature plot was calcu-lated to determine the exothermic heat generatedduring polymerisation.13 The rate of polymerisationwas determined by calculating the gradient of thelinear portion of the temperature-versus-time plot, atthe tset.

27 Additionally, the cumulative thermal necrosisindex (TNI) was determined at two temperature levels:�44�C and �55�C.10–13,27

Polymerisation reaction and degree of monomerconversion

The extent of the polymerisation reaction was moni-tored using a DSC 6 differential scanning calorimeter(Perkin Elmer Inc., USA) and analysis was conductedusing Pyris Manager and Data Analysis software(Perkin Elmer Inc., USA). Each test was conducted

under isothermal conditions; with the specimen heldat 22� 1�C for 45min.

The time at which polymerisation began (tinitial), andfinished (tfinal) was calculated. Energy released duringpolymerisation (�Hexp), monomer conversion fraction,xt, and total monomer conversion of each specimen, x,were determined. In addition to monitoring the poly-merisation reaction of the PMMA bone cement, DSCwas also used to calculate the reaction constant, k.28,29

Microscopical analysis

Raman spectroscopy. Raman spectroscopy was con-ducted on the fractured surface of each mechanicaltest specimen using a Raman Station Fusion R1,fitted with an 8200 Detector Element Echelle CCDdetector (Avalon Instruments, UK). The resultant spec-tra were analysed with the peaks of interest being atRaman shifts of 1616 cm�1 (G-Band) and 1320 cm�1

(D-Band). The shift in the peak position of theG-band was examined with the compressive strain onthe MWCNT-COOH exerted by the PMMA matrixbeing estimated.9

Confocal Raman spectroscopy. Confocal Raman spectros-copy was conducted on the fractured surface of eachtest specimen post mechanical testing using theRenishaw inVIA Reflex spectrometer connected to aLeica microscope and equipped with a 514 nm laserand a lateral spatial resolution of 2 mm. This techniquewas utilised to attain a surface spectral Raman map anda depth profile Raman map in order to characterise theextent of MWCNT dispersion within the PMMA bonecement. A total of 120 spectra covering a Enger-printregion of 20 mm2 were recorded. Three maps each con-taining 120 spectra from three different random selectedlocations were collected for each sample. Depth profileRaman maps were also conducted at the same locationson the specimen as the surface spectral Raman map,which involved increasing the laser exposure (20 s). Foreach specimen, three maps each containing five spectrawere collected. The resultant spectra were analysed withthe peaks of interest being the phonon modes located at1616 cm�1 (G-Band) and 1320 cm�1 (D-Band).

Data pre-processing

It is accepted that spectral artefacts can have a consid-erable effect on the interpretation of Raman data.Therefore, it is necessary to distinguish between chem-ical information and undesired effects. Before any datainterpretation can be performed, it is important thatpre-processing of data is conducted. To this end,Origin 7 software (OriginLab, USA) was used to rou-tinely pre-process the raw data. Initially, band

Ormsby et al. 211

at UNIVERSITE LAVAL on July 16, 2014jba.sagepub.comDownloaded from

alignment was performed to correct for instrumentalspectral shifts, consequently all spectra werenormalized using the area of the 1449 cm�1 band. Amulti-point baseline correction and smoothing by afive-point Savitzky–Golay Elter was also employed foreach Raman spectrum. Contributions of the fused silicasubstrate were subtracted, however small contributionsmay still be present. The area of individual vibrationalbands was computed using the curve Etting option ofOrigin 7 software.

Field emission scanning electron microscopy

Field emission scanning electron microscopy(FE-SEM) analysis of the fractured surfaces of eachbone cement combination following mechanical testingwas conducted using a JEOL 6500 FEG SEM(Advanced Microbe am, Inc., USA) with operatingvoltages of 5.0 kV. Specimens were mounted on alumin-ium discs using a cold cure resin (Extec Corp, Enfield,CT, USA) and allowed to cure for 24� 1 h. The speci-mens were subsequently sputtered with gold prior toSEM examination.

Statistical analysis

For each property determined, the results were evalu-ated for statistical significance using a one-way analysisof variance (one-way ANOVA), with p< 0.05 denotingsignificance. Post-hoc tests were conducted using theStudent-Newman-Keuls and Duncan methods (SAS8.02; SAS Institute, USA).

Results

Mechanical characterisation

Static properties. The static mechanical properties ofSimplex PTM bone cement with 0.1wt% MWCNT-COOH added and the control cement are summarisedin Table 1. The compressive strength and compressive

modulus of the MWCNT-COOH/PMMA bone cementexhibited marginal improvements (p< 0.1) of 2% and1%. However, significant improvements (p< 0.001) inbending strength (42%) and bending modulus (55%)were observed. Similarly, significant improvements(p< 0.001) in the fracture toughness (22%) when com-pared to the control PMMA bone cement were seen. Itis interesting to observe that the significant improve-ments in bending strength and modulus did not correl-ate to enhancements of the same magnitude forcompressive strength and modulus of the PMMAcement tested. It is suggested that the methods adoptedfor specimen preparation, specimen configuration anddifferent modes of loading employed during these dif-ferent tests could account for this. It has previouslybeen reported that different loading regimes evaluatediffering reinforcement mechanisms within the speci-men microstructure, therefore dissimilar responses areexpected.30,31 Wagner and Chu32 also reported distinc-tions in mechanical properties when testing three dentalcore ceramic-based materials. They found significantdifferences in the biaxial flexural strength, but reportedno significant difference for the indentation fracturetoughness or compressive properties for the materialstested.

Fatigue properties. Examining the fatigue test results, sig-nificant improvements were observed when 0.1wt%MWCNT-COOH was incorporated into the SimplexPTM bone cement. The mean and median number ofcycles to failure increased (p� 0.001) by 64% and66% when compared with the control cement(Table 2). The fatigue life estimates were determinedusing the probability of fracture method (Table 2).14

Incorporating 0.1wt% MWCNT-COOH to SimplexPTM bone cement significantly improved (p� 0.001)the Weibull minimum fatigue life (N0) and thenumber of cycles at which 50% of specimens failed(N50) values by 80% and 70%. Adding MWCNT-COOH to the Simplex PTM cement significantlyincreased (p� 0.001) the Weibull characteristic fatigue

Table 1. Mechanical (static) properties (mean� SD) for control and carboxyl functionalised multi-walled carbon

nanotube (MWCNT-COOH)/Simplex PTM bone cements. Percentage difference between control and MWCNT-COOH/

Simplex PTM bone cements for mechanical (static) property is also indicated.

Mechanical properties: static Control MWCNT-COOH/Simplex PTM Change (%)

Compressive strength (MPa) 102.38� 5.38 104.47� 6.05 2.00

Compressive modulus (MPa) 3659� 468 3672� 356 0.35

Bending strength (MPa) 59.68� 13.31 102.88� 19.26a 41.99

Bending modulus (MPa) 2731� 743 6120� 1701a 55.38

Fracture toughness (MPa m1/2) 2.40� 0.63 3.07� 0.39a 21.73

ap Value less than 0.001.

212 Journal of Biomaterials Applications 29(2)

at UNIVERSITE LAVAL on July 16, 2014jba.sagepub.comDownloaded from

life (Na) value by 64%, and a reduction in the Weibullslope (b) value of 14% was also recorded. The additionof MWCNT-COOH to the Simplex PTM bone cementsignificantly improved (p� 0.001) the fatigue perform-ance index (I) by 52%.

Applying Weibull theory, the probability of survival(Ps) for a given number of cycles was determined forthe Simplex PTM bone cement and MWCNT-COOH/Simplex PTM cement (Figure 1). For a given Ps of 0.3,the Weibull life was determined for Simplex PTM bonecement as 100,000 cycles. In contrast, incorporating0.1wt% of MWCNT-COOH into the Simplex PTM

bone cement increased the Weibull life to 150,000cycles for the same Ps level. Figure 1 indicates theMWCNT-COOH/Simplex PTM bone cement exhibiteda similar starting Ps value as the control cement, whichincreased through the low (0–50,000), intermediate(50,000–200,000) and higher (200,000–1,000,000)

number of cycles. This clearly indicates that incorpor-ating 0.1wt% MWCNT-COOH into Simplex PTM

bone cement demonstrates a significantly higher Pslevel when compared to the control bone cement.

Thermal characterisation

Setting properties. Incorporating 0.1wt% MWCNT-COOH into the Simplex PTM bone cement significantly(p< 0.001) reduced the extent of the polymerisationreaction (Table 3). Significant reductions (p< 0.001)of 8% in the maximum temperature (Tmax) were deter-mined for the MWCNT-COOH/Simplex PTM bonecement when compared to the control. Incorporating0.1wt% MWCNT-COOH also significantly increased(p< 0.001) the setting time (tset) for the Simplex PTM

bone cement by 17%. Adding the MWCNT-COOHreduced (p< 0.1) the rate of the polymerisation reaction

1

0.9

0.8

0.7

0.6

0.5

0.4

0.3

0.2

0.1

01000

Sur

viva

l pro

babi

lity

10000 100000

Number of cycles to failure (Nf)

Simplex P

Simplex P + MWCNT-COOH

1000000 10000000

Figure 1. Survival probability vs. the log of the number of cycles to failure for the control and 0.1 wt% MWCNT-COOH/Simplex

PTM bone cements.

Table 2. Mechanical (fatigue) properties (mean� SD) for control and carboxyl functionalised multi-walled carbon nanotube

(MWCNT-COOH)/Simplex PTM bone cements. Percentage difference between control and MWCNT-COOH/Simplex PTM bone

cements for each mechanical (fatigue) and Weibull parameter is also indicated.

Mechanical properties: fatigue Control MWCNT-COOH/Simplex PTM Change (%)

Mean number of cycles to failure (cycles) 83,648� 33219 137,025� 46737a 64.00

Median number of cycles to failure (cycles) 79,337� 33219 131,462� 46737a 66.01

Weibull minimum fatigue life (N0) (cycles) 42,313 76,218a 80.13

50% probability of fracture life (N50) (cycles) 79,433 134,896a 69.82

Weibull characteristic fatigue life (Na) (cycles) 87,202 143,197a 64.21

Fatigue performance index (I) 82,303 125,401a 52.37

Weibull slope (b) 0.89 0.77a 13.91

ap Value less than 0.001.

Ormsby et al. 213

at UNIVERSITE LAVAL on July 16, 2014jba.sagepub.comDownloaded from

by 4%. The exotherm generated during polymerisationwas significantly reduced (p< 0.001) on addition of0.1wt% MWCNT-COOH by 16% when compared tothe control cement. The inclusion of 0.1wt%MWCNT-COOH to Simplex PTM bone cement signifi-cantly reduced (p< 0.001) the TNI values at >44�C(28%) and >55�C by 28%.

Polymerisation reaction and degree of monomerconversion

Incorporation of 0.1wt% MWCNT-COOH signifi-cantly influenced (p< 0.001) the polymerisation reac-tion of the Simplex PTM bone cement (Table 4).Increases in the onset of polymerisation (tonset), thetime at which the reaction had completed (tf) and dur-ation of the polymerisation (�t) were measured at 58%,24% and 52% when compared with the control PMMAbone cement. The peak energy and time at which thispeak energy (tpeak) occurred significantly increased(p< 0.001) by 672% and 12%, respectively. The peakarea of the polymerisation (�Hexp) also increased(p< 0.001) by 116% demonstrating that the inclusionof the 0.1wt% MWCNT-COOH had a significant role

in reducing the rate of polymerisation. This was con-firmed when the reaction rate constants k1 and k2 werecalculated, both of which significantly decreased(p< 0.001) by 21% and 15%, respectively. An insignifi-cant reduction (p> 0.05) in the percentage monomerconversion was recorded (8%) when the MWCNT-COOH was incorporated into the Simplex PTM bonecement.

Microscopical analysis

Raman spectroscopy. Typical Raman spectra for the con-trol and MWCNT-COOH/Simplex PTM bone cementsare observed in Figure 2. From these spectra, the ratioof the normalised intensities of the G-Band andD-Band (IG/ID) and the compressive strain were calcu-lated. MWCNT-COOH/Simplex PTM bone cementshowed an increase in the IG/ID ratio when comparedwith the pure MWCNT-COOH. Examining thetheoretical compressive strain exerted on theMWCNT-COOH within the Simplex PTM matrix itwas determined that a compressive strain of –1.72%was exerted on the MWCNT-COOH by the SimplexPTM cement matrix. Typical surface Raman and

Table 4. Thermal properties (mean� SD) for control and carboxyl functionalised multi-walled carbon nanotube (MWCNT-COOH)/

Simplex PTM bone cements as determined using the DSC technique. Percentage difference between control and MWCNT-COOH/

Simplex PTM bone cements for each property is also indicated.

Thermal properties: DSC Control MWCNT-COOH/Simplex PTM Percentage difference (%)

Polymerisation onset (tonset) (min) 6.74� 0.83 10.63� 7.05a 58

Polymerisation end (tfinal) (min) 12.01� 3.18 14.94� 8.06 24

Polymerisation duration (�t) (min) 5.27� 0.86 8.04� 7.03 52

Peak energy height time (tpeak) (min) 5.62� 6.49 6.28� 11.08 12

Peak energy (mW) 2.19� 0.85 16.94� 21.07 672

Peak area (�Hexp) (mJ) 503.47� 223.69 2296.05� 2407.58a 116

k1 (min�1) 0.0185� 0.0003 0.0147� 0.0018a�21

k2 (min�1) 0.3955� 0.0032 0.3346� 0.0872a�15

Monomer conversion (%) 68.73� 8.06 63.35� 5.51 �8

ap Value less than 0.001.

Table 3. Thermal properties (mean� SD) for control and carboxyl functionalised multi-walled carbon nanotube (MWCNT-COOH)/

Simplex PTM bone cements as determined using ISO5833:2002 test protocol. Percentage difference between control and MWCNT-

COOH/Simplex PTM bone cements for each thermal property is also indicated.

Thermal Properties: ISO5833 Control MWCNT-COOH/Simplex PTM Change (%)

Maximum temperature (Tmax) (�C) 76.62� 4.79 70.50� 6.87a�8.00

Exotherm generated during cure (�C min�1) 1686.30� 41.41 1449.72� 126.54a�16.32

Setting time (tset) (min) 9.62� 0.32 11.57� 0.279a 16.86

Rate of polymerisation (�C min�1) 114.20� 18.92 109.35� 16.76b�4.44

TNI� 44�C (s) 107.43� 20.40 84.28� 4.03a�27.47

ap Value less than 0.001.bp Value less than 0.01.

214 Journal of Biomaterials Applications 29(2)

at UNIVERSITE LAVAL on July 16, 2014jba.sagepub.comDownloaded from

depth profile Raman maps for the MWCNT-COOH/Simplex PTM bone cement are shown in Figures 3and 4. The dispersion of the MWCNT-COOH withinthe PMMA bone cement was determined to be homo-genous as the intensity of the G-Band and D-Band forthe surface Raman and depth profile Raman maps were31� 2% and 30� 10%, respectively.

Discussion

Within this study, it has been demonstrated that incor-porating a relatively low level of carboxyl functiona-lised MWCNT significantly improved the static andfatigue mechanical properties of a proprietary PMMAbone cement by reducing the crack initiation and sub-sequent propagation within the cement mantle. It ishypothesised that these improvements in mechanicalperformance are due to the homogenous dispersion ofMWCNT within the Simplex PTM cement matrix. It isbelieved that this dispersion was amplified by the nega-tively charged carboxyl groups attached to theMWCNT backbone. This hypothesis is supported bythe uniform MWCNT distribution as indicated by thesurface map and depth profile of the bone cement usingRaman Spectroscopy.

The presence of well-dispersed MWCNT-COOHwithin the cement matrix reduced the occurrence ofagglomerates forming. The incidence of agglomerationcan potentially act as sites of stress concentration, andtherefore locations of crack initiation, which leads to

premature mechanical failure.3,10,13 Consequently, thehomogenous dispersion demonstrated would also facili-tate a greater level of chemical interaction between thecarboxyl groups and the Simplex PTM bone cementmatrix, potentially resulting in successful transfer ofstress during physical loading. Singh et al. have previ-ously demonstrated the successful capability for in vivoimplantation of a biomimetic CNT-hydroxyapatite-polymethyl methacrylate nanocomposite.33 The successof their approach was largely down to the ability totailor the physical and chemical properties of the com-posite for the crucial requirements of biological integra-tion. This was achieved via controlling the surfacemodification design, which had a significant influ-ence on the extent of the interactions between CNTand the hydroxyapatite-polymethyl methacrylate.Fractographic analysis of the surfaces of the bonecement post-mechanical failure showed evidence ofMWCNT pull-out from the PMMA bone cementmatrix (Figure 5).

This expected mode of failure would occur as theinterfacial shear strength between the MWCNT-COOH and the bone cement matrix is less than thetensile strength of the MWCNT-COOH (0.2TPa),3

this trend was also noted in our previous studies.10,12,13

An accumulation of material is visible at the end of apulled-out MWCNT (Figure 6), which further supportsthe hypothesis that a degree of chemical interactionexists between the PMMA bone cement andMWCNT-COOH, again, this is in agreement with our

D-Band G-Band

MWCNT-COOH

MWCNT-COOH-Simplex PTM

Control

275022501750Raman wavenumber (cm–1)

Nor

mal

ised

inte

nsity

(a.

u.)

1250750250

Figure 2. Raman spectra for the MWCNT-COOH, MWCNT-COOH/Simplex PTM and control bone cements, the D-band and

G-bands have been highlighted. The laser wavelength used for microscopical Raman was 785 nm.

Ormsby et al. 215

at UNIVERSITE LAVAL on July 16, 2014jba.sagepub.comDownloaded from

previous work.3 It is postulated that this accumulatedmaterial may be due to the elastic recoil of a tensilestrained region of the polymer matrix.

From a theoretical perspective, the chemical inter-action between the PMMA constituent of the SimplexPTM bone cement and the MWCNT-COOH wouldcause a polymer sheath to form around MWCNT-COOH during polymerisation. Once tensile failurehas occurred, this polymeric sheath would be energet-ically released from one end of the MWCNT-COOHand be ‘‘coiled-up’’ on the opposing MWCNT-COOHtip. This would be a direct result of a chemical inter-action or ‘‘chemical wetting’’ between MWCNT-COOH and the polymeric cement matrix. It is

postulated that this chemical interaction between theMWCNT-COOH and the polymeric cement matrixwould facilitate mechanical augmentation of the bonecement. These findings are corroborated by Marrset al.3 and Ding et al.,34 both research groups alsoobserved polymeric matrix material on the end ofMWCNT post fatigue failure. The exact mechanismof reinforcement is unclear – either typical fibrereinforcement (e.g. fibre pull-out) or chemical inter-action and therefore the transfer of load is most dom-inant. However, it can be inferred that each mechanismplays an active role in enhancing the mechanical per-formance of PMMA bone cement. The improvementsin the static and fatigue mechanical properties are

120

100

80

60

8000

6000

4000

2000

8000

6000

4000

2000

2000

Raman shift / cm–1

Raman shift / cm

–1

500 10001500 2000 2500 3000

1000 3000

40

20

0

120

Dat

a se

t

Cou

nts

Dat

a se

t

3×10ˆ4

2×10ˆ4

1×10ˆ4

100

80

60

40

20

0

Figure 3. Representative 2D and 3D Raman confocal map for the surface of the MWCNT-COOH/Simplex PTM bone cement. The

laser wavelength used for confocal Raman was 514 nm.

216 Journal of Biomaterials Applications 29(2)

at UNIVERSITE LAVAL on July 16, 2014jba.sagepub.comDownloaded from

deemed significant as the mechanical failure of PMMAbone cement has been cited as the main cause for asep-tic loosening of prosthetic implants.1

Previously, Marrs et al.3 investigated the influence ofunfunctionalised MWCNT on mechanical properties ofPMMA-based bone cements. The authors reportedmodest (13%–24%) and significant (>300%) improve-ments in bending and fatigue properties of methylmethacrylate-styrene bone cement at a loading levelof 0.2wt% MWCNT. However, we have previouslyhighlighted that these studies did not use clinically rele-vant methods to incorporate the MWCNT into thecement and nonstandardised test methods for specimenpreparation.3 Therefore, our previous studies focusedon incorporating MWCNT into the bone cementmatrix using proprietary bone cement mixing

10

8

6

3000

2000

2500

1500

1000

40000

30000

20000

10000

2000

Raman shift / cm–1

Raman shift / cm–1

500 1000 1500 2000 2500 3000

1000 3000

4

2

0

10

Data

set

Cou

nts

Dat

a se

t

3×10ˆ4

4×10ˆ4

2×10ˆ4

1×10ˆ4

86

42

0

Figure 4. Representative 2D and 3D Raman confocal depth profiles for the MWCNT-COOH/Simplex PTM bone cement. The laser

wavelenght used for confocal Raman was 514 nm.

Figure 5. SEM images illustrating examples of MWCNT pull-

out from the PMMA bone cement matrix.

Ormsby et al. 217

at UNIVERSITE LAVAL on July 16, 2014jba.sagepub.comDownloaded from

technology.10,12,13 These studies used in-house designedPMMA bone cement that was analogous to a commer-cially available PMMA bone cement, however thiscement is not typically used during joint replacementsurgery. Therefore, it was essential that the effect ofincorporating MWCNT-COOH into proprietary bonecement (Simplex PTM) was also characterised. It is pro-posed that the chemical interaction between theSimplex PTM bone cement and MWCNT-COOHoccurs via two different mechanisms. Firstly, theCOOH groups on the MWCNT-COOH could poten-tially share electrons with and therefore (in theory)covalently bond to the polymer chains within theSimplex PTM bone cement. This would occur at eachof the hydrogen positions on the polymer chain via anesterification reaction (Figure 7).

In addition, the benzene ring of the styreneco-polymer found within the Simplex PTM will alsointeract with the side wall of the MWCNT-COOHthrough p-p stacking (Figure 8). Through this mechan-ism of p-p stacking, a stacked arrangement of aromaticmolecules (e.g. carbon) will overlap p-orbital’s andthus, share electron charges. If either or both of thesechemical interactions occurred, there would be a

chemical bond generated between the host PMMAbone cement and the MWCNT-COOH. This wouldallow for greater mechanical and thermal propertiesof the bone cement, a trend which is evident in thiscurrent study.

In terms of the Raman spectroscopy data, it is sug-gested that the increase in the disorder-induced D-band(IG/ID ratio) further supports the hypothesis of chem-ical interactions between the COOH functional groupson the MWCNT-COOH and the -OH group of theMMA liquid monomer, as well as the styrene-PMMAco-polymer powder. The addition of the MWCNT-COOH to the liquid monomer may therefore increasethe number of defect sites present on the MWCNT-COOH due to these chemical interactions. This indica-tion of chemical interactions between the MWCNT-COOH and the MMA liquid monomer is further evi-denced by the compressive strain exerted on theMWCNT-COOH by the Simplex PTM bone cementmatrix. The compressive strain values measured inthis study for the COOH functionalised MWCNT areapproximately 10% higher than that reported byMcClory et al.9 They reported compressive strains of�1.56% for a polyurethane composite containing0.1wt% MWCNT. It is postulated that this improve-ment in compressive strain is due to the effective chem-ical interaction between the MWCNT-COOH and theSimplex PTM bone cement. If the bone cement is indeedchemically grafting to the MWCNT-COOH at thefunctional COOH groups, then there would be anexpected increase in surface interactions and a corres-ponding increase in mechanical properties for the

Figure 6. SEM images showing of the PMMA cement matrix

forming a sheath around isolated MWCNT and consequently

developing into ‘‘coiled-up’’ polymer on the tip of the MWCNT

due to elastic recoil (a) low magnification and (b) high

magnification.

H

CC C

H

C

O

O

n

H

C

H

H H

C

H

H H

H H

H H

H H

Figure 7. Schematic representation showing the repeating unit

of the Simplex PTM bone cement with the hydrogen molecules

highlighted. Each of these hydrogen molecules provides a position

whereby the MWCNT-COOH could potentially form a chemical

bond.

218 Journal of Biomaterials Applications 29(2)

at UNIVERSITE LAVAL on July 16, 2014jba.sagepub.comDownloaded from

MWCNT-COOH/Simplex PTM bone cement, asobserved in this study. If the results presented hereare contextualised with previous studies,10,12,13 it isnoteworthy that the mechanical properties are greateron addition of 0.1wt% MWCNT-COOH to theSimplex PTM bone cement when compared with theincorporation of MWCNT-COOH at the same loadinglevel to the previously used model PMMA bonecement.10,12,13 It is suggested that this is due to theincreased chemical interaction between the MWCNT-COOH and the polymer chains of the Simplex PTM

bone cement because of the benzene ring of the styreneco-polymer. In contrast the model PMMA bone cement(utilised in our previous studies) does not contain abenzene ring, and as a consequence will demonstrateless chemical interaction with the MWCNT-COOH. It is worth mentioning that the compressivestrain reported in this study is 21% higher than thatpreviously reported for the MWCNT/PMMA bonecement composite based on the modelformulation.10,12,13

The reduction in the exothermic polymerisationreaction of Simplex PTM bone cement reported in thisstudy could potentially decrease the development ofresidual stresses within the bone cement mantle as aconsequence of thermal shrinkage.27 Furthermore, themaximum temperatures recorded during polymerisa-tion (>100�C) can result in a permanent cessation of

blood flow, whilst causing necrosis of the bonesurrounding the implant. The results from this studydemonstrate that incorporation of a low level ofMWCNT-COOH assisted in the dissipation of heatproduced during the exothermic polymerisation reac-tion of Simplex PTM bone cement. The MWCNT-COOH used here exhibit a thermal conductivity of>3000Wm�1k�1 (Nanocyl data sheet), it is thereforeproposed that the MWCNT-COOH acted as a heatsink within the Simplex PTM bone cement and, thus,assisted in dissipation of heat generated during thepolymerisation reaction. The inclusion of MWCNT-COOH led to an altered rate of polymerisation forthe Simplex PTM bone cement. A slower rate of poly-merisation extended the tset for the bone cement to fullypolymerise, which in turn reduced the Tmax and TNIvalues. It is suggested that the presence of MWCNT-COOH not only altered the kinetics of thepolymerisation reaction but also played an active rolein dissipation of the heat energy generated. This theorywas supported by the DSC analysis completed in thisstudy. The addition of MWCNT-COOH extended thepolymerisation reaction, as well as the time at which thepeak polymerisation occurred. The reaction rate con-stants (k1 and k2) significantly decreased (p< 0.001) inaddition to the percentage monomer conversion alsoreducing, although not significantly. It is suggestedthat this may have be due to the MWCNT-COOH

(B)

(A)

(D)

(C)

MWCNT C

OH

CH2

CH3

CH3

C

CCC

C

C

C

C

C

C

CCC

CC

CC

C

C

C

CC

C

CCC

C

C

C

C C

C

CC

C

C

C

C

C

C

CC

C C

CC

C

H

O

O

n

O

Figure 8. Schematic representation showing: (A) MWCNT-COOH; (B) wall of the MWCNT-COOH indicating the hexagonal

arrangement of carbon atoms; (C) repeating unit of the Simplex PTM bone cement; (D) p–p stacking that occurs between the

hexagonal carbon atom arrangement within the wall of the MWCNT-COOH and the benzene ring of the repeating unit of Simplex

PTM bone cement.

Ormsby et al. 219

at UNIVERSITE LAVAL on July 16, 2014jba.sagepub.comDownloaded from

acting as radical scavengers during the radical polymer-isation of the MMA, therefore hindering its normalroute. This hypothesis is supported by a recent studyby Goncalves et al.,14 where these workers suggestedthat CNT can act as radical scavengers during polymer-isation. These findings would also support the hypoth-esis that the addition of the MWCNT-COOHsignificantly influenced the polymerisation reactionchemically, which resulted in significant reductions inextent of the exothermic reaction.

Work is currently on-going to investigate the extentof the chemical interaction occurring between theMWCNT-COOH and PMMA bone cement. Briefly,the MWCNT-COOH PMMA cement will be dissolvedin chloroform, following an extensive purificationphase the recovered MWCNT-COOH will be analysedusing X-ray photoelectron spectroscopy and Fouriertransform infrared spectroscopy in an effort to verifyand quantify the incidence of PMMA attached viacovalent bonding.

The findings reported within this study may be con-sidered significant as cement mantle failure (due to fati-gue related cracking) is consistently cited as one of theleading contributor to failure of cemented total jointreplacements. Additionally the elevated temperaturesexperienced in vivo due to the highly exothermic poly-merisation reaction of the PMMA bone cement canlead to thermally induced bone necrosis, which inturn can lead to aseptic loosening of the implant, andsubsequent implant failure.1–3

Conclusions

This study has extended the findings of our previouswork to augment the mechanical and thermal proper-ties of commercially available Simplex PTM bonecement, which is routinely used for joint replacementsurgical procedures.10,13 Additionally, Raman spectros-copy has been used to explain the mechanism ofreinforcement of PMMA bone cement withMWCNT. DSC was also used to examine the effectof MWCNT-COOH on the polymerisation reactionof the Simplex PTM cement. The addition ofMWCNT-COOH (0.1wt%) to the Simplex PTM bonecement improved the mechanical properties. Thehomogeneous dispersion of MWCNT-COOH withinthe Simplex PTM cement delayed crack propagationthrough the cement mantle during static and fatigueloading. The addition of MWCNT-COOH also signifi-cantly reduced the exothermic polymerisation reactionof the Simplex PTM bone cement. These reductions inexotherm were attributed to the MWCNT-COOHacting as heat sinks within the bone cement matrix.Raman spectroscopy was employed to demonstratethat a chemical interaction between the

MWCNT-COOH and the Simplex PTM bone cementoccurred. This interaction allowed for chemical bond-ing of the MWCNT-COOH to the Simplex PTM bonecement, thereby facilitating the successful transfer ofmechanical load. Homogenous dispersion of theMWCNT-COOH within the bone cement was demon-strated by Raman mapping.

Acknowledgements

The authors thank Nanocyl S.A., Belgium, and Lucite

International Ltd., UK, for supplying the MWCNT-COOHand PMMA bone cement. The authors would like to acknow-ledge the support of the Tyndall National Institute, Republicof Ireland. This support was provided through the Science

Foundation Ireland-funded National Access Programme(Project NAP349).

Conflict of interest

None declared.

Funding

This research was financially supported by the Department ofEducation and Learning, Northern Ireland and the RoyalAcademy of Engineering through the Royal Academy of

Engineering/Leverhulme Senior Research FellowshipProgram.

References

1. Kuehn KD, Ege W and Gopp U. Acrylic bone cements:composition and properties. Orthop Clin North Am 2005;36: 17–28.

2. Meyer PR, Lautenschlager EP and Moore BK. On the

setting properties of acrylic bone cement. J Bone JointSurg 1973; 55: 149–156.

3. Marrs B, Andrews R, Rantell T, et al. Augmentation of

acrylic bone cement with multiwall carbon nanotubes.J Biomed Mater Res Part A 2006; 77: 269–276.

4. Harper EJ, German MJ, Bonfield W, et al. A comparison

of VersaBondTM, a modified acrylic bone cement withimproved handling properties, to Palacos R, Simplex P.In: Trans sixth world biomater cong, Kamuela (Big

Island), 2000, p.540.5. Choi ES, Brooks JS, Eaton DL, et al. Enhancement of

thermal and electrical properties of carbon nanotube poly-mer composites by magnetic field processing. J App Phys

2003; 94: 6034–6039.6. Xie XL, Mai YW and Zhou XP. Dispersion and alignment

of carbon nanotubes in polymer matrix: A review. Mater

Sci Eng R 2005; 49: 89–112.7. Andrews R and Weisenberger MC. Carbon nanotube

polymer composites. Curr Opin Solid State Matter 2004;

8: 31–37.8. Spiegelberg SH and McKinley GH. Characterization of

the curing process of bone cement with multi-harmonic

shear rheometry. In: Trans 24th Ann Mtg, Soc Biomater,San Diego, CA. 1998; pp.283.

220 Journal of Biomaterials Applications 29(2)

at UNIVERSITE LAVAL on July 16, 2014jba.sagepub.comDownloaded from

9. McClory C, McNally T, Baxendale M, et al. Electricaland rheological percolation of PMMA/MWCNT nano-composites as a function of CNT geometry and function-

ality. Eur Poly J 2010; 46: 854–868.10. Ormsby R, McNally T, Mitchell CA, et al. Incorporation

of multiwalled carbon nanotubes to acrylic based bonecements: Effects on mechanical and thermal properties.

J Mech Behav Biomed Mater 2010; 3: 136–145.11. Ormsby R, McNally T, O’Hare P, et al. Fatigue and bio-

compatibility properties of a poly(methyl methacrylate)

bonecement with multi-walled carbon nanotubes. ActaBiomaterialia 2011; 8: 1201–1212.

12. Ormsby R, McNally T, Mitchell CA, et al. Influence of

multiwall carbon nanotube functionality and loading onmechanical properties of PMMA/MWCNT bonecements. J Mater Sci Mater Med 2010; 21: 2287–2292.

13. Ormsby R, McNally T, Mitchell CA, et al. Thermal andrheological properties of PMMA/MWCNT bonecements. Carbon 2011; 49: 2893–2904.

14. Goncalves G, Cruz SMA, Ramalho A, et al. Graphene

oxide versus functionalized carbon nanotubes as reinfor-cing agent in a PMMA/HA bone cement. Nanoscale2012; 4: 2937–2945.

15. British Standard for Implants for Surgery: Acrylic ResinCements. BS ISO 5833:2002.

16. Dunne NJ, Orr JF, Mushipe MT, et al. The relationship

between porosity and fatigue characteristics of bonecements. Biomaterials 2003; 24: 239–245.

17. Hastings NAJ and Peacock JB. Statistical distributions.London, UK: Butterworth, 1975.

18. Weibull W. Fatigue testing and analysis of results.New York, USA: Macmillian, 1961.

19. Shigley JE and Mischke CR. Mechanical engineering

design, 5th ed. New York: McGraw-Hill, 1989.20. Murphy BP and Prendergast PJ. On the magnitude and

variability of the fatigue strength of acrylic bone cement.

Int J Fatigue 2000; 22: 855–864.21. Britton JC, McInnes P, Ledoux WR, et al. Shear bond

strength of ceramic orthodontic brackets to enamel. Am J

Orthod Dentofac 1990; 98: 348–353.

22. Topoleski LDT, Ducheyne P and Cuckler JM. A fracto-graphic analysis of in vivo poly(methyl methacrylate)bone-cement failure mechanisms. J Biomed Mater Res

1993; 24: 135–154.23. Davies JP, O’Connor DO, Greer JA, et al. Comparison

of the mechanical properties of Simplex P, ZimmerRegular and LVC bone cements. J Biomed Mater Res

1987; 21: 719–730.24. Davies JP, O’Connor DO, Greer JA, et al. Influence of

antibiotic impregnation on the fatigue life of Simplex P

and Palacos R acrylic bone cement, with and withoutcentrifugation. J Biomed Mater Res 1989; 23: 379–397.

25. Davies JP and Harris WH. The effect of addition of

methylene blue on the fatigue strength of Simplex Pbone cement. J Appl Biomater 1992; 3: 81–85.

26. Lewis G and Austin GE. Mechanical properties of

vacuum mixed acrylic bone cement. J Appl Biomater1994; 5: 307–314.

27. Dunne N and Orr J. Curing characteristics of acrylicbone cement. J Mater Sci Mater Med 2002; 13: 17–22.

28. Barrett KEJ. Determination the rates of thermal decom-position of polymerisation initiators with a differentialscanning calorimeter. J Appl Polym Sci 1976; 11:

1617–1626.29. Filipovic J, Petrovic-Djakov D, Vrhovac LJ, et al. DSC

investigation of isothermal bulk copolymerization of

alkylaryl methacrylates. J Therm Anal and Calorim1993; 40: 735–740.

30. Lewis G and Mladsi S. Correlation between impactstrength and fracture toughness of PMMA-based bone

cements. Biomaterials 2000; 21: 775–781.31. Wagner W and Chu T. Biaxial flexural strength and

indentation fracture toughness of three new dental core

ceramics. J Prosthet Dentist 1996; 76: 140–144.32. Singh MK, Gracio J, LeDuc P, et al. Integrated biomim-

etic carbon nanotube composites for In Vivo systems.

Nanoscale 2010; 2: 2855–2863.33. Ding W, Eitan A, Fisher FT, et al. Direct observation of

polymer sheathing in carbon nanotube-polycarbonate

composites. Nano Letters 2003; 3: 1593–1597.

Ormsby et al. 221

at UNIVERSITE LAVAL on July 16, 2014jba.sagepub.comDownloaded from