Microsoft Word - Azizi Proof Dec 4.docx   Vol  2,  No  1  (2014)   ISSN  2167-­‐8677  (online)   DOI  10.5195/d3000/2014.24         http://dentistry3000.pitt.edu     New  articles  in  this  journal  are  licensed  under  a  Creative  Commons  Attribution  4.0  United  States  License.     This  journal  is  published  by  the  University  Library  System,  University  of  Pittsburgh  as  part  of  its  D-­‐Scribe  Digital  Publishing  Program  and  is  cosponored   by  the  University  of  Pittsburgh  Press.   A  gas  adsorption  porosimetry  analysis  of  Portland  cement  prepared   by  compaction  vs.  compaction  with  indirect  ultrasonic  agitation   Nader  Azizi1,  Yltze  P.  Cubas2,  Maria  F.  Orellana3     1DDS,  MSc,  Private  Prac.ce,  New  York  City,  USA   2DDS.  Post-­‐doctoral  fellow,  Department  of  Orofacial  Sciences,  School  of  Den4stry,  University  of  California,  San  Francisco,  USA   3DDS,  MSc,  PhD.  Director  Predoctoral  Orthodon4c  Program,  Department  of  Orofacial  Sciences,  Division  of  Orthodon4cs,  School  of  Den&stry,  University  of  California,   San  Francisco,  USA Abstract     Background:  Mineral   trioxide   aggregate   (MTA),   is   commonly   used   in   endodon9c   and   re-­‐ stora&ve  procedures.  Objec&ve:  Our  objec&ve  was  to  introduce  gas  adsorp&on  porosimetry   as  a  viable  method  for  evalua1on  of  general  porosity  and  specific  pore  characteris1cs  of  set   Portland  cement  used  in  subs1tute  for  MTA,  to  inves1gate  the  effect  of  two  different  obtu-­‐ ra#on  methods  (compac#on  and  compac#on   in  conjunc#on  with   indirect  ultrasonic  agita-­‐ !on  of  the  cement  paste),  and  to  evaluate  the  correla.on  between  the  specific  pore  char-­‐ acteris(cs   to   compressive   strength   in   general.  Material   and   Methods:   Portland   cement   samples  were  prepared  and  divided  into  two  groups  based  on  compac4on  techniques.  An   ini#al   stereomicroscopic   evalua*on  was   done   to   assess   any   differences   in   appearance   of   pores  randomly  selected  from  either  of  the  two  experimental  groups.  Specific  pore  charac-­‐ teris&cs  and  compressive  strength  werequan&fied  by  a  gas  adsorp&on  porosimeter  and  an   Instron  universal  tes(ng  machine.  A  two-­‐tailed  student  t-­‐test  was  used  for  sta,s,cal  com-­‐ parison   of   data,   and   a   regression   analysis  was   done   to   evaluate   the   correla6on  between   each  specific  pore  characteris-c  and  compressive  strength  in  general.  Results:  The  gas  po-­‐ rosimetry  method   provided  measurable   values   rela2ng   to   specific   pore   characteris2cs   of   Portland   cement.   The   stereomicroscopy   evalua6on   revealed  marked  differences  between   samples  from  the  two  groups,  namely  visibly  larger  pores  both  on  the  outside  surface  and   in  cross-­‐sec$ons  of  specimens  prepared  by  the  indirect  ultrasonic  ac$va$on  method.  Con-­‐ clusions:   Gas   adsorp*on  porosimetry   is   a   feasible  method   for   evalua*on  of   specific   pore   characteris)cs  of  Portland  cement  and  poten)ally  other  dental  materials  as  well.       Cita%on:  Azizi  N,  Cubas  Y,  Orellana  M.  (2014)  A  gas  ad-­‐ sorp%on  porosimetry   analysis   of   Portland   cement  pre-­‐ pared   by   compac,on   vs.   compac,on   with   indirect   ul-­‐ trasonic   agita+on.   Den+stry   3000.   1:a001   doi:10.5195/d3000.2014.24   Received:  March  25,  2014   Accepted:  October  15,  2014   Published:  November  3,  2014   Copyright:   ©2014   Azizi   et   al.   This   is   an   open   access   ar#cle   licensed  under  a  Crea#ve  Commons  A2ribu#on   Work  4.0  United  States  License.   Email:  OrellanaM@den,stry.ucsf.edu       Introduc!on     The   key   requirement   for   the   strength  and  stability  of  hydraulic  cement  is   that   the   hydrates   formed   on   immediate   reaction   with   water   should   be   essentially   insoluble  in  water.    Mineral  Trioxide  Aggre-­‐ gate  (MTA)  is  an  inorganic  hydraulic  cement   used   as   a   restorative  material   in   dentistry.     Portland  cement  is  reported  to  be  the  main   component   of   this   inorganic   hydraulic   ce-­‐ ment.   Similar   to   other   inorganic   cements,   MTA   can   set   and   harden   in  moist   environ-­‐ ments   [1],   a   property  which  makes   it   ideal   for   use   in   dentistry.     In   addition,   MTA   has   been   shown   to   provide   a   predictable   seal   against  leakage  in  bacterial  and  dye  leakage   studies   [2-­‐4].     The   calcium  hydroxide   com-­‐ ponent   released   during   its   hydration   reac-­‐ tion   provides   alkalinity   to   this   material,   which  may  account  for  both  its  antibacterial   and   hard-­‐tissue   growth   induction   proper-­‐ ties  [5,6]     MTA   placement   technique   affects   some  of  the  structural  and  material  proper-­‐ ties  of   the   set   cement   [7,8].     These  proper-­‐ ties   range   from   the   cement’s   adaptation   to   its   container’s   walls,   to   changes   in   its   re-­‐ sistance  to  leakage  and  variations  in  its  sur-­‐ face   and   internal   porosity.   It   is   now   well   accepted   that   pores   decrease   the  mechani-­‐ cal  properties  of  cement  [9].    Therefore,  any   technique   that   would   potentially   affect   the   porosity   of   MTA   may   similarly   affect   its   mechanical  properties.  Selection  of  handling   and  placement  methods,  which  optimize  the   physical   and   mechanical   properties   of   the   cement,  may  prove  bene/icial.    For  instance,   commonly   observed   .lexural   weakness   of   cements   has   been   related   to   presence   of   large  voids  in  a  set  material.  The  removal  of   macro-­‐defects  during  preparation  has  been   shown   to   increase   the   )lexural   strength   of   hydraulic  cements  [10].    Similar  effects  may   potentially   be   expected   in   respect   to   com-­‐ pressive  strength.     Pores   in   hydraulic   cements,   such   as   MTA,   may   stem   from   a   number   of   sources.     They   could   be   either   internal   in   the  source,  such  as  those  created  as  a  result   of  the  material’s  hydration  reaction,  or  they   could   have   an   external   source.     Internal   sources   of   porosity   may   have   a   profound   impact   on   its   mechanical   properties   as      A  gas  adsorp)on  porosimetry  analysis  of  Portland  cement  prepared  by  compac)on  vs.  compac)on  with  indirect  ultrasonic  agita)on   Vol  2,  No  1  (2014)        DOI  10.5195/d3000/2014.24    http://dentistry3000.pitt.edu   2   Figure  1:    Best  &it  lines  through  intercepts  of  speci/ic  pore  characteristics  and  compressive  strength  of  samples.    a,b.  Graphs  show  a  slight   positive  correlation.  c.  Graph  shows  a  slight  negative  correlation.     powder-­‐to-­‐water   is   the   most   important   parameter   in*luencing   the   porosity   of   hy-­‐ draulic   cements   [11].   Selection  of   the  pow-­‐ der-­‐to-­‐water  ratio  would  affect  the  porosity   of  a  cement  mix  due  to  its  effect  on  the  size   and  distribution  of  its  capillary-­‐pores  which   are   inevitably   formed  during   the  hydration   reactions.     Similarly,   external   sources   of   porosity  may  also  in-luence  some  structural   properties   of   restorative   cements.     For   ex-­‐ ample,   direct   ultrasonic   activation   of   an   MTA  cement  paste  during  its  placement  has   been   linked   to   entrapment   of   larger   sized   air-­‐bubbles   in   the   set   specimen   [12]   possi-­‐ bly  due  to  a  proposed  “whipping”  motion  of   the  ultrasonic  delivery  device.    These  larger   pores,  similar  to  those  affecting  the  (lexural   strength   of   the  material,   may   affect   on   the   compressive  strength  of  the  material.         Although  MTA  is  traditionally  used   in   low   pressure   bearing   areas,   its   use   in   high  pressure  bearing  areas  may  potentially   be   indicated   (i.e.   as   bases   under   other   re-­‐ storative  dental  materials).    Some  have  pro-­‐ posed   the  use  of  Portland  cement  as  an   in-­‐ expensive  alternative   to  other  more  expen-­‐ sive  dental  restorative  materials  for  restora-­‐ tion   of   coronal   defects   [13],   with   potential   uses   in   economically   challenged   areas.     When   used   in   such   critical   locations,   both   MTA   and   Portland   cement   must   withstand   compressive   forces   that   could   challenge   their   clinical   longevity   through   time.   Other   factors   affecting   porosity   such   as   pore   size   distributions,  microcracks,  interface,  among   others   are   also   important   in   determining   the  mechanical   properties   of   cement  mate-­‐ rials  [14,  15].    However,  porosity  which  can   be   semi-­‐empirically   and   concisely   used   to   describe   the   relationship   between   strength   and   microstructure   of   porous   material   is   still   being   studied   [16,   17].   Porosity   has   been   shown   to   display   an   inverse   relation-­‐ ship   to   compressive   strength,   with   reduc-­‐ tion  of  porosity  leading  to  increases  in  com-­‐ pressive  strength  of  a  solid  but  porous  ma-­‐ terial.       Gas  adsorption  has  been  one  of  the   most  popular  techniques  used  for  the  study   of   pore   structure   in   materials   that   contain   micropores  and  mesopores,  or  mainly  pores   in   cement   pastes   with   radii   between   1nm   and   approximately   60nm.     Therefore,   gas   adsorption  technique  can  mainly  character-­‐ ize   gel   pores   and   small   and  medium   capil-­‐ lary   pores.   Gas   adsorption   methods   are   based   on   measurements   of   the   amount   of   gas  adsorbed  on  the  surface  of  a  powder  on   the   monomolecular   depth   of   the   particle’s   surface   [18].     When   a   porous   solid   is   ex-­‐ posed   to   gas   of   a   certain   volume  and  pres-­‐ sure,   it   begins   to   adsorb   the   gas  molecules   on   its   outside   surface   and   inside   its   pores.     The   amount   of   gas   adsorbed  during   an   ad-­‐ sorption  experiment  can  then  be  calculated   by   either   a   gravimeric,   volumetric,   or   ther-­‐ mal   conductivity   measurement   method.     From   the   amount   of   gas   adsorbed   and   the   corresponding   relative   pressure   recorded   during   the   experiment,   a   great   number   of   methods   are   developed   for   analysis   which   include   methods   for   determining   the   pore   volume,   pore   size   distribution,   and   the   Brunauer-­‐Emmett-­‐Teller   (BET)   speci.ic   surface   area   of   the   material.   Surface   area   measurements   are   the   most   widely   used   means  for  characterization  of  porous  mate-­‐ rials   .   Since   the   surface   corresponds   to   the   roughness   of   the   particle   and   its   porous   interior,   gas   adsorption   is   the   preferred   technique  [19,20].     We   investigated   the   effect   of   two   common   clinically   practiced   placement   techniques   on   porosity   and   compressive   strength   of   Portland   cement   in   an   in   vitro   model.    We   compared   the   porosity   of   sam-­‐ ples   of   Portland   cement   placed   by   either   compaction  alone   (Group  1)  or   compaction   in  conjunction  with   indirect  ultrasonic  acti-­‐ vation  of  the  cement  paste  during  its  place-­‐ ment  (Group  2).    Initially,  stereomicroscopic   inspections   of   samples   were   made   in   an   attempt   to   make   visual   comparisons   be-­‐ tween  specimens  selected  from  either  of  the   two   experimental   groups.   Later,   a   gas   ad-­‐ sorption  porosity  method  was  used  to  quan-­‐ tify   and   compare   few   speci1ic   pore   charac-­‐ teristics   of   the   cement   samples   –   namely   their   BET   speci-ic   pore   surface   area,   pore   volume,   and   pore   size   distribution-­‐   from   both   groups.   Compressive   strengths   of   samples   were   quanti/ied   using   an   Instron   universal  material  testing  machine.         Materials  and  Methods     In  a  clean  glass  dappen  dish,  1.0  g   of  Portland  cement  was  mixed  with  0.32  ml   of   deionized   water   using   a   metal   mixing   spatula.     Care   was   taken   to   gently  mix   the   two   components   to   avoid   introduction   of   large   air   bubbles   into   the   cement   paste.     Mixing   was   done   for   30   seconds   until   the   material   took   a   consistency   which   allowed   for   easy   transfer   and   placement   of   its   ali-­‐ quots   by   an   amalgam   carrier   into   metallic   moulds.   Then   each   sample   followed   treat-­‐ ment  by  either  of  the  two  methods  of  obtu-­‐ ration.  In  Group  1,  8  samples  were  obturat-­‐ ed   into  the  molds  using  a   ten  second  appli-­‐ cation   of   an   equal   and   constant   amount   of   vertical   force   to   the   cement   paste.     A   3ive   pound  weight  placed  on  top  of  a  piston  con-­‐ nected   to   the   metallic   obturator   delivered   the   vertical   component   of   the   obturation   force.    In  Group  2,  8  samples  were  prepared   similarly,   with   the   difference   of   indirect   ultrasonic   energy.     The   tip   of   a   vibrating   ultrasonic   instrument   (BUC-­‐2)   was   placed   along  an  area  about  one   inch  above   the   tip   of   the   obturator   to   transfer   a   10   second   burst   of   ultrasonic   energy   to   the   metallic   instrument.     The   obturator,   in   turn,   propa-­‐ gated   the   ultrasonic   energy   to   the   paste   along  with  delivering  its  vertical  component   force  of  compaction.  A  gridded  plastic  sepa-­‐ rator  was  placed  over  a  thin  pool  of  water  at   the  bottom  of   the  container  with  the  molds   placed   on   top.   The   plastic   separator   was   used   to   prevent   the   samples   from   directly   coming  in  contact  with  the  water  reservoir,   while   allowing   them   to   set   at   nearly   100%   a b ca b c    A  gas  adsorp)on  porosimetry  analysis  of  Portland  cement  prepared  by  compac)on  vs.  compac)on  with  indirect  ultrasonic  agita)on   Vol  2,  No  1  (2014)        DOI  10.5195/d3000/2014.24    http://dentistry3000.pitt.edu   3   Figure  2:  Samples  of  set  cement  placed  side  by  side  un-­‐ der  5x    magni&ication.    Note  the  presence  of  visible  pores   on   the   surface   of   specimen   from   the   indirect   ultrasonic   activation  group  (A);    Slight  variation  in  color  marks  the   border   between   aliquots   of   cement   placed   to   make   up   Figure  3:  At  10x  magni,ication,  large  internal  pores  (air-­‐bubbles)   could  be  seen   in   cross   sections  of  samples   from   the   indirect  ul-­‐ trasonic  activation  group  (A);  Transition  areas  between  aliquots   of  cement  present  as  “watered-­‐down,  weak  fault-­‐lines.     humidity   and   at   room   temperature.     After   one   week,   setting   of   samples   was   visually   con$irmed.     Samples  were   then  pushed   out   of  each  mold  and  prepared  for  porosity  and   compressive  strength  evaluations.    Random-­‐ ly   selected   samples   from   each   group   were   prepared   for   stereomicroscopic   evaluation   (5X,  10X,  and  20X  magni,ications.)    The  ex-­‐ ternal   surfaces   of   samples   were   initially   evaluated   for   presence   of   visibly   large   sur-­‐ face   pores.     Then,   samples   were   ground   down  to  make  half-­‐cylinders  of  samples  and   exposing   their   internal   appearance.     Stereomicroscopic   evaluations   of   internal   surfaces   were   also   done   for   presence   of   notably  visible  pores.   Gas  adsorption  porosimetry  evaluation    Samples   were   sent   to   Micrometrics   (Mi-­‐ cromeritics,  Norfolk,  GA,  USA.)  and  evaluat-­‐ ed   after   varying   durations   of   setting   time.     Specimens   were   de-­‐gassed   at   40°C   for   16   hours   then   placed   in   a   sample   tube   and   heated  under   vacuum  or   .lowing   gas   to   re-­‐ move   contaminants   on   the   surfaces   of   the   samples.   The   sample   tube  was   then   placed   in   the   analysis   port   of   a   2420   Accelerated   Area  and  Porosimetry  System  for  automatic   analysis.   The   krypton   adsorption   isotherm   was  recorded  at  120  K.         Compressive  strength  analysis      Cylindrical   samples   from   each   experimental   group   were   carefully   placed   on  their  vertical  axes  to  stand  perpendicular   to   the   two   horizontal   plates   of   the   Instron   testing   machine.     The   mobile   upper   plate   was   set   to   advance   at   a   uniform   speed   of   1mm   per   second   to   deliver   the   crushing   vertical   forces   to   the   samples.     Increasing   vertical   forces  were  applied  to  each  sample   until   structural   failure   was   achieved,   at   which  point   this  maximum  weight  was  rec-­‐ orded.   Statistical  Analsysis     Statistical   analysis   comparing   the   average  values  of   speci/ic  pore   characteris-­‐ tics   and   compressive   strength   of   samples   was  done  using  a  two-­‐tailed  student’s  t-­‐test.     Correlation   and   regression   tests  were  used   to   look   for   any   relationship   between   each   speci&ic   pore   characteristic   (BET   speci,ic   surface   area   of   pores,   pore   volume,   and   pore   size   distribution)   and   compressive   strength  in  general.     Results     In   Table   1,   we   ob-­‐ served   an   apparent   trend   in   BET   pore   surface   area   and   a   trend   in   pore   volume   of   sam-­‐ ples   in  Group  1,   however;   the   difference   in   values   are   not   statistically  signi#icant.     Additionally,   there   is   an   ap-­‐ parent  trend  in  pore  size  and  a   trend   in  compressive  strength   of  samples  in  Group  2,  but  the   difference  in  values  is  not  sta-­‐ tistically  signi(icant.   Regression   and   Correla-­‐ tion  Analysis   Compressive   strength   vs.   BET   speci!ic   pore  surface     This   revealed   a   positive   regression   with   a   correlation   coef!icient   of   +0.65.   This   shows   that   as   the   BET   pore   surface   area   of  samples  increased,  so  did   the  compressive  strength  of   the  samples  (Figure  1a).   Compressive   strength   vs.   Pore  volume     A   positive   regres-­‐ sion   with   a   correlation   coef!icient  of  +0.64  shows   that   if   the  pore  volume  of   samples   increased,   so   did   the   compressive   strength   of   the   samples   (Figure   1b).     Compressive   strength   vs.   Pore   size   dis-­‐ tribution     A  negative  regression  with  a  corre-­‐ lation   coef!icient   of   -­‐0.42   shows   that   if   the   pore   size   of   samples   increased,   the   com-­‐ pressive   strength   of   samples   tends   to   de-­‐ crease  (Figure  1c).   Stereomicroscopy     5X   magni!ications   of   randomly   selected   samples   from   each   group   were   placed   side   by   side.     Initially   we   noticed   visible   differences   in   the   surface   appear-­‐ ance   of   the   two   cylindrical   specimens.   The   sample   prepared   by   vertical   compaction   force   alone   presented   a   more   uniform   ap-­‐ pearance  of  its  outside  surface  area  with  no   obvious   large  porosities   or   transition   lines.     However,  this  specimen  presents  numerous   larger  porosities  and  a  number  of  transition   lines  which   correspond   to   the   junction   be-­‐ tween  aliquots  of  cement  paste  place  on  top   of  each  other  during  the  segmental  compac-­‐ tion  of  the  cement.     The  same  side  by  side  comparison   of  cross  sections  of   samples  were  arranged   as   before,   but   at   10X  magni+ication   reveals   similar   arrangement   of   larger   pore   sizes   and  transition   lines  present   throughout   the   interior   bulk   of   the   sample   prepared   by   compaction   in   conjunction   with   ultrasonic   activation  (Figure  3a).  A  closer  look  at  tran-­‐ sition   lines  visualized   at   25X  magni2ication   reveals   weak   faults   created   between   ali-­‐ quots   of   cement   samples   placed   by   the   combined  compaction  and  ultrasonic  activa-­‐ tion  method  (Figure  3b).     Discussion     In   this   study,   we   introduced   gas   adsorption   porosimetry   as   an   acceptable   method   for   evaluation   and   distinction   of   speci&ic   pore   characteristics   of   a   dental   re-­‐ storative   material,   namely   its   BET   speci4ic   surface   area,   pore   volume   and   pore   size   distribution.     This   method   of   analysis   gen-­‐ erates   quantitative   data   with   values   that   re#lect  on  those  speci#ic  pore  characteristics   of  solid  but  porous  materials.    According  to      A  gas  adsorp)on  porosimetry  analysis  of  Portland  cement  prepared  by  compac)on  vs.  compac)on  with  indirect  ultrasonic  agita)on   Vol  2,  No  1  (2014)        DOI  10.5195/d3000/2014.24    http://dentistry3000.pitt.edu   4   Ying-­‐zi  et  al   “gas  adsorption  porosimetry   is   suitable   for   the   materials   with   apertures   smaller   than  50nm  or   the  powder  with   the   particle  size  smaller  than  100nm  [20].”  Our   data  showed  that  gas  adsorption  porosime-­‐ try   could   be   used   for   comparative   evalua-­‐ tion   and   statistical   analysis   of   pore   charac-­‐ teristics  of  different  dental  materials   in  ad-­‐ dition   to   evaluating   the   effects   of   chemical   or  physical  manipulations  on  the  materials’   speci&ic  pore  characteristics.  Lee  at  al  stated   that   gas   adsorption   technique   gives   the   largest   amount   of   information   for   the   po-­‐ rous  structure  of  solids  [21].     When  comparing   the  pore  charac-­‐ teristics   of   samples   placed   by   either   of   the   two   tested   placement  methods,   our   overall   preliminary  results   show  no  signi2icant  dif-­‐ ferences   between   the  BET   pore   surface   ar-­‐ ea,  pore  volume,   and  pore   size  distribution   of   samples   from   either   of   the   two   groups   however;   these  data  displayed  a  number  of   observed   trends   which   indicate   a   positive   effect   on   compaction   in   conjunction   with   indirect  ultrasonic  technique.    The  numbers   of  specimens   tested   in   this  study  were   lim-­‐ ited   due   to   the   high   cost   of   a   larger   scale   analysis.     Based   on   the   limited   number   of   samples  tested,  eight  samples  per  group,  no   statistically   signi+icant   differences   in   either   the  average  BET  speci0ic   surface  area,  pore   volume  or  pore  size  distribution  of  samples   from   either   of   the   two   groups   may   be   re-­‐ ported.    However,  from  this  data  trends  may   be  observed.    First,  our  data  showed  a  trend   toward  larger  average  surface  area  of  pores   in   those   samples   prepared   by   compaction   alone.    In  contrast,  there  was  a  trend  toward   smaller   average   surface   area   of   pores   in   those   samples   prepared   by   compaction   in   conjunction  with  ultrasonic  activation  of  the   paste   during   its   placement.     Second,   we   observed   similar   results  with   samples   pre-­‐ pared  by  compaction  alone  showing  a  trend   toward   larger   average   pore   volumes   com-­‐ pared   to   those   prepared   by   the   indirect   ultrasonic   activation   method.     Lastly,   this   data  revealed  a  trend  which  showed  smaller   average   pore   size   distributions   in   samples   prepared  by  compaction  compared  to  those   made   using   the   indirect   ultrasonic   activa-­‐ tion  method.         Surface   and   cross   sectional   stere-­‐ omicroscopic   inspection   of   the   samples   from  each  group   led  us   to  a  number  of   ini-­‐ tial   impressions  related  to  structural  differ-­‐ ences  between   the   two.     It   appeared   that   a   sample  prepared  by  compaction  alone  had  a   more   uniform   outer   surface     which   was   devoid   of   any   noticeably   large   sized   air-­‐ entrapped   pores   that   were   more   readily   observed  on  the  outer  surface  of  the  sample   prepared   by   the   indirect   ultrasonic   activa-­‐ tion   of   the   paste   during   its   placement.     A   similar   pattern   was   also   observed   in   mid-­‐ line  cross  sectional  preparations  of  samples   from   both   groups.     These   cross   sectional   preparations   also   revealed   the   presence   of   more   numerous   air-­‐entrapped   pores   in   samples   prepared   by   ultrasonic   activation.     Similar  observations  have  also  been  report-­‐ ed   in   other   investigations  where   a   “direct”   activation   of   the   cement   through   insertion   of  the  vibrating  ultrasonic  tip  into  the  paste   is   thought   to   create   a   “whip-­‐ ping”  motion  in  the  paste,  lead-­‐ ing  to  introduction  of  larger  air   bubbles   into   the   paste   [14].   Another   explanation   for   this   observation  may   be   related   to   a  phenomenon  of   the  action  of   ultrasonically   vibrating   in-­‐ struments   in   liquid  media,   call   “acoustic   cavitation   [22]”.   This   phenomenon   may   potentially   introduce  larger  bubbles  into  a   more  viscous  medium,   such   as   that   of   our   Portland   cement   paste  specimens.         Another   noteworthy   observation   that   was   made   only  in  the  sample  prepared  by   the   indirect   ultrasonic   activa-­‐ tion  was   the   presence   of   hori-­‐ zontal   demarcation   lines   be-­‐ tween   aliquots   of   cement   placed  during  obturation  of  the   mold.    A   plausible   explanation   for   formation  of  such  demarcations  may  be   that  the  agitation  provided  by  the  ultrasonic   energy  may  have  potentially  forced  the  solid   particles   of   Portland   cement   to   stack-­‐up   more   closely   against   each   other   by   forcing   any   unbound   water   out   of   the   spaces   be-­‐ tween  those  particles,  and  essentially,  caus-­‐ ing  sedimentation  of  the  solid  entities  in  the   water-­‐cement   paste  mixture.   Other   reports   have   shown   that   a   critical   increase   in   the   liquid   component   of   hydraulic   cement   would   lead   to  a  decrease   in   the  strength  of   the  material.     These   “watered-­‐down”   inter-­‐ faces  may  ultimately  present  as  those  weak   “fault-­‐lines”   seen   in   the   indirect   ultrasoni-­‐ cally   prepared   samples   in   our   magni!ied   images.       The   presence   of   visually   larger   pores   in   samples   placed  by   the   indirect   ul-­‐ trasonic  activation  method  concur  with  our   report   of   larger   values   for   average   pore   sizes   along   with   the   expected   correspond-­‐ ingly   smaller   average   pore   volume   and   av-­‐ erage   pore   surface   area   of   samples   pre-­‐ pared  with   the  aid  of  an   indirect  ultrasonic   activation   method.     Such   .indings   are   in   contrast   to  possibly  more   sedimentation  of   solid  Portland  cement  particles  in  the  paste   prepared  by  the  indirect  ultrasonic  method.     This  conclusion  was  based  on  the  pooling  of   water   observed   on   top   of   each   aliquot   of   cement   paste   placed   to   make   up   a   whole   sample.    This  would  suggest  that  sa   mples   prepared   by   the   indirect   ultrasonic   activation   method   would   expectedly   have   smaller  pores  sizes,  while  presenting  larger   pore   surface   areas   and   pore   volumes.     Yet   this  was  not   supported  by  our  quantitative   data.    We  believe  that  the  indirect  ultrasonic   activation   creates   more   densely   packed   specimens   with   smaller   pores   and   poten-­‐ tially   larger   pore   surface   areas   and   pore   volumes.    The   larger   air  bubbles,   as   visual-­‐ ized   in   stereomicroscopic   evaluations,  may   be   outliers   that   skew   the   data   enough   to   support   results   reporting   larger   average   pore  size  distributions  and  smaller  average   pore   surface   areas   and   pore   volumes   that   contradict   the   visually   observed   signs   of   sedimentation   in   the   indirectly   activated   ultrasonic  method.    Excluding  these  outliers   we  report  that  samples  prepared  by  indirect   ultrasonic   activation   are   indeed   more   densely   packed   at   a   more   microstructural   level   when   compared   to   those   placed   by   compaction  alone.    Such  extrapolations  may   be   further   supported   when   comparisons   between   data   re*lecting   the   average   com-­‐ pressive   strength   of   samples   from   the   two   groups  is  made.     Our   data   showed   that   the   place-­‐ ment   method   did   not   have   a   statistically   signi%icant   effect   on   the   average   compres-­‐ sive   strength   of   the   samples   from   the   two   groups;  however,  there  was  a  trend  toward   samples   in   the   ultrasonically   activated   group  having  a  higher  average  compressive   strength   values   than   those   samples   placed   Table  1:  Average  values  of  group  1  and  group  2.     Group  1   Group  2   p-­‐value   BET  Speci!-­‐ ic  Pore  Sur-­‐ face   6.175m2/g   SD=3.0   5.172m2/g   SD=4.072   0.762   Pore  Vol-­‐ ume   0.0257m3/g   SD=0.02   0.0241m3/g   SD=4.072   0.922   Pore  Size   Distribu-­‐ tion   239.886  Å   SD=144.27   318.09  Å   SD=162.27   0.4984   Compres-­‐ sive   Strength   88.667kg   SD=24.8   92.0kg     SD=  34.7   0.894      A  gas  adsorp)on  porosimetry  analysis  of  Portland  cement  prepared  by  compac)on  vs.  compac)on  with  indirect  ultrasonic  agita.on   Vol  2,  No  1  (2014)        DOI  10.5195/d3000/2014.24    http://dentistry3000.pitt.edu   5   by   compaction   alone.     Presence   of   few   air-­‐ bubbles  may   be   compensated   by   the  more   densely  packed  cement  particles  at  a  micro-­‐ structural   level   in   the   indirect   ultrasonic   method.    This  density  may  be  related  to  the   observed   higher   average   compressive   strength  of  samples  from  this  group.     Conclusions     Gas   adsorption   porosimetry   is   a   method   that   can   be   potentially   used   for   evaluation  of  speci/ic  pore  characteristics  of   dental   restorative   materials.     Larger   scale   studies  are  needed  to  evaluate  the  superior-­‐ ity   of   one   method   of   placement   over   the   other.     References     1. Evaluation  of  setting  properties  and   retention  characteristics  of  mineral   trioxide  aggregate  when  used  as  a   furcation  perforation  repair  mate-­‐ rial.  Sluyk  SR,  Moon  PC,  Hartwell  GR.   J  Endod.  1998  Nov;24(11):768-­‐71.   PMID:  9855831.     2. A  comparison  of  MTA,  Super-­‐EBA,   composite  and  amalgam  as  root-­‐end   filling  materials  using  a  bacterial  mi-­‐ croleakage  model.  Adamo,  HL,  Bu-­‐ ruiana  R,  Schertzer  L,  Boyland  RJ.  In-­‐ tEndod  J.  1999  May;32(3):197-­‐203.   PMID:  10530207.     3. Leakage  evaluation  of  root  end  fill-­‐ ing  materials  using  endotoxin.  Tang   HM,  Torabinejad  M,  Kettering  JD.  J   Endod.  2002  Jan;28(1):5-­‐7.  PMID:   11806652.     4. Dye  leakage  of  four  root  end  filling   materials:  effects  of  blood  contami-­‐ nation.  Torabinejad  M,  Higa  RK,   McKendry  DJ,  Pitt  Ford  TR.J  Endod.   1994  Apr;20(4):159-­‐63.  PMID:   8035153.     5. Antibacterialeffects  of  some   rootendfillingmaterials.  Torabinejad   M,  Hong  CU,  Pitt  Ford  TR,  Kettering   JD.  J  Endod.  1995  Aug;21(8):403-­‐6.   PMID:  7595152.     6. Reaction  of  rat  connective  tissue  to   implanted  dentin  tubes  filled  with  a   white  mineral  trioxide  aggregate.   Holland  R1,  Souza  Vd,  Nery  MJ,  Far-­‐ acoJúnior  IM,  Bernabé  PF,  Otoboni-­‐ Filho  JA,  DezanJúnior  E.  Braz  Dent  J.   2002;13(1):23-­‐6.  PMID:  11870957.     7. A  comparison  of  laterallycon-­‐ densedgutta-­‐percha,  thermoplasti-­‐ cizedgutta-­‐percha,  and  mineral  tri-­‐ oxide  aggregate  as  root  canal  filling   materials.  Bogen  G.  J  Endod.  2004   Dec;30(12):826.  PMID:  15564859.     8. A  quantitative  comparison  of  the  fill   density  of  MTA  produced  by  two   placement  techniques.  Yeung  P1,   Liewehr  FR,  Moon  PC.  J  Endod.  2006   May;32(5):456-­‐9.  PMID:  16631848.     9. Porosity-­‐dependence  of  effec   10. tive  mechanical  properties  of  pore-­‐ solid  composite  materials.  Pal  R.  J   Com-­‐pos  Mater.  2005   Jun;39(13):1147-­‐1158.     11. Flexural  strength  and  porosity  of  ce-­‐ ments.  Birchall  J,  Howard  AJ,  Kendall   K.  Nature.  1981  Jan;289:388-­‐390     12. Bye  GC  (1999).  Portland  Cement,   2ndedition:  Composition,  Production   and  Properties.  United  King-­‐ dom:Thomas  Telford  Ltd.248  p.     13. Placement  of  mineral  trioxide  aggre-­‐ gate  using  two  different  techniques.   Aminoshariae  A,  Hartwell  GR,  Moon   PC.  J  Endod.  2003  Oct;29(10):679-­‐ 82.  PMID:  14606796.     14. Comparative  chemical  study  of  MTA   and  Portland  cements.  Oliveira  MG,   Xavier  CB,  Demarco  FF,  Pinheiro  AL,   Costa  AT,  Pozza  DH.  Braz  Dent  J.   2007  18(1):3-­‐7.  PMID:  17639192.     15. Microstructure  and  strength  of  hy-­‐ drated  cement.Feldman  RF,   Beaudoin  JJ.  CemConcr  Res.  1976   May;6(3):398-­‐400.     16. Mehta  PK  (1993).  Concrete:  Struc-­‐ ture,  Properties,  and  Materials.  New   York:  Prentice  Hall.  704  p.     17. Relationship  between  pore  struc-­‐ ture  and  mechanical  properties  of   ordinary  concrete  under  bending  fa-­‐ tigue.Zhang  B.  Cem  and  Concr  Res.   1998  May;28(5):699–711.     18. Compressive  strength  and  pore   structure  of  high-­‐performance  con-­‐ crete  after  exposure  to  high  temper-­‐ ature  up  to  800C.Chan  Y,  Luo  X,  Sun   W.  CemConcr  Res.  2000  30(2):247-­‐ 251.     19. Asthana  R,  Kumar  A,  Dahotre  N   (2006).  Materials  Processing  and-­‐ Manufacturing  Science,  First  Edition.   United  Kingdom:  Butterworth-­‐ Heinemann.  656  p.     20. Aligizaki  KK  (2005).  Pore  Structure  of   Cement-­‐Based  Materials:  Testing,   Interpretation  and  Requirements.   Boca  Raton:  CRC  Press,  LLC.  432  p.     21. Determination  of  pore  size  distribu-­‐ tion  and  surface  area  of  several  ma-­‐ terials  using  mercury  porosimetry   and  gas  adsorption.  Tian  Y,  Chen  Ke-­‐ fu.  China  Pulp  and  Paper.  2004-­‐04.     22. The  Characterization  of  Porous  Sol-­‐ ids  from  Gas  Adsorption  Measure-­‐ ments.Lee  CK,  Chiang  AST,  Tsay  CS.   Key  Eng  Mater.  1996  115:21-­‐44.     23. Ultrasonic  debridement  of  root  ca-­‐ nals:  acoustic  cavitation  and  its  rele-­‐ vance.  1988.  Ahmad  M,  Pitt  Ford  TR,   Crum  LA,  Walton  AJ.  IntEndod  J.   2009  May;42(5):391-­‐8.  PMID:   19356172.