Summary
Objectives.  The  goals  of  this  study  were  to  investigate  the  phase transformation and microstructure of dental porcelain bodies densified via a moving laser beam and to develop an understanding of how the microstructure of the dental porcelain varies with the laser processing condition and the position relative to the center of the laser beam.
Methods. A moving laser beam was used to scan and densify a commercial dental porcelain  powder  bed.  The  porcelain  powder  compact  was  also  sintered  using  a furnace at different temperatures. The phase transformation and microstructure of these  dental  porcelain  bodies  were  compared  and  investigated  using  a  host  of analytical  instruments  including  scanning  electron  microscopy,  X-ray  diffraction,thermogravimetric analysis, and quantitative image analysis.
Results. Based on the temperature dependence of the leucite content in the dental porcelain derived from the furnace-sintered samples, the temperature distribution in  the  dental  porcelain  body  during  laser  densification  was  established.  The microstructure of the laser-densified porcelain body was found to be dependent of the  location  relative  to  the  center  of  the  laser  beam  and  the  average  laser densification temperature.Significance. The understanding of how the microstructure of dental porcelain bodies varies with the laser processing condition and the location with respect to the center of the laser beam was developed. Based on this understanding, the laser processing condition has been optimized to achieve the desired microstructure and densification of dental porcelain bodies simultaneously.
Q2004 Academy of Dental Materials. Published by Elsevier Ltd. All rights reserved.

KEYWORDS
Dental porcelain;Laser densification;Phase transformation;Microstructure;Solid freeform,fabrication

Introduction
The dominant procedure currently used for perma- nent  fixed  prosthodontics  is  porcelain-fused-to- metal  (PFM)  restoration [1–5] .  There  are  several steps involved in PFM restoration, including casting of a dental alloy base from a wax pattern of the patient’s  teeth  and  building  multiple  layers  of dental porcelains on the cast alloy base by several furnace firing processes. As such, PFM restoration is a  time  consuming  and  labor  intensive  work  with labor costs accounting for about 90% of the final cost to the patient [5] . In order to reduce costs and provide faster and better service to dental patients, a  new  dental  restoration  method  multi-material laser densification (MMLD), based on the principle of solid freeform fabrication (SFF), is being devel- oped [6–9] .
SFF is an automated manufacturing process that builds  three-dimensional  complex-shaped  struc- tures layer-by-layer directly from CAD data source. More than 24 SFF techniques have been developed by  research  organizations  and  commercial  com- panies  in the recent past [10–12] , most of which scan an energy-delivery beam, e.g. a light, laser or electron beam in a computer-controlled pattern to densify  a  liquid  or  solid  powder  layer-by-layer  in order to fabricate physical object directly from a computer model. In the MMLD process, both dental alloy and porcelain powders are delivered line-by- line or layer-by-layer, and subsequently densified by  a  laser  beam,  thereby  the  name  of  MMLD. Through this automated process, an artificial tooth is  expected  to  be  fabricated  from  a  computer model  without  part-specific  tooling  and  human intervention  in  an  hour  or  two,  and  therefore MMLD offers the potential to substantially reduce the labor cost and increase the dental restoration rate.
In the MMLD process, it is expected that there exists  a  temperature  gradient  within  the  area irradiated by a laser beam because of the Gaussian distribution  of  the  output  power  intensity  of  the laser beam. Thus, the temperature at the center of a laser-scanned path can be much higher than the average temperature of the sampling area, which is read by an emissivity-measuring infrared pyrometer that  provides  the  temperature  reading  based  on integration of the thermal radiation emitted by the sampling  area.  Temperature  gradients  are  also present  in  the  entire  laser-densified  body  as densification is achieved by a moving laser source. Because of these temperature gradients, the phase transformation  and  microstructure  of  a  laser- densified  dental  porcelain  body  may  vary  from location to location. Furthermore, the microstruc- ture and phases obtained may not be the desired ones  because  of  the  non-uniform  temperature distribution. These issues are especially important for  dental  porcelain  powders  currently  used  in dental  laboratories  if  they  are  utilized  for  the MMLD  process.  Most  of  the  dental  porcelain powders  used  today  consist  of  two  phases:  the feldspar  glass  matrix  and  crystalline  leucite  par- ticles (K 2 O $ Al 2 O 3 $ 4SiO 2 ) [1–5] . The feldspar glass, which has a coefficient of thermal expansion (CTE) of 7–8 ! 10 K 6 / 8 C, relies on the presence of leucite particles,  a  high  CTE  phase  (20–25 ! 10 K 6 / 8 C),  to raise its CTE to match the CTE of the dental alloy (typically 13–16 ! 10 K 6 / 8 C). Without the match in CTE  between  the  dental  porcelain  and  alloy, cracking  in  porcelain  develops  during  the  PFM fabrication  process [2,13] .  Similarly,  cracking  is expected to occur in the MMLD process if the CTE mismatch is not minimized or eliminated. There- fore,  there  is  a  critical  need  to  investigate  the phase transformation and microstructure of dental porcelain bodies densified via a moving laser beam. To meet this critical need, the aims of this study are to: (i) investigate the phase transformation and microstructure of dental porcelain bodies densified via  a  moving  laser  beam;  (ii)  develop  an  under- standing of how the microstructure of the dental porcelain varies with the laser processing condition and the location with respect to the center of the laser  beam;  and  (iii)  based  on  the  understanding developed, optimize the laser processing condition to achieve the desired microstructure and densifi- cation of dental porcelain bodies simultaneously.



Materials and methods The  dental  porcelain  powder  Ceramco q II  Silver Body  (batch  number  99081250)  was  provided  by Dentsply  Ceramco,  Burlington,  NJ.  The  chemical composition of the porcelain is confidential; how- ever, it is within 5% of the nominal composition of the Component No. 1 in Weinstein patent, which is SiO 2 —63.40%,  Al 2 O 3 —16.70%,  K 2 O—14.19%,  Na 2 O— 3.41%, CaO—1.50%, and MgO—0.80% (wt%) [13] . The as-received  porcelain  powder  had  angular  shapes and their equivalent particle sizes ranged from 5 to 50 m m.  Furthermore,  the  powder  contained  two phases, the feldspar glass matrix and the crystalline leucite particles. Laser  densification  of  the  dental  porcelain powder  was  conducted  using  an  integrated  SFF system, a schematic of which is shown in Fig. 1 .In the laser densification process a continuous wave laser beam from a 50 W CO 2 laser was steered into a process chamber with a controlled atmosphere by a series of motion-controlled mirrors. The scanning and sintering of the desired area of a powder bed within the process chamber was achieved via an x – y table controlled by a computer. The entire densi- fication  process  was  closed-loop-temperature controlled  using  a  pyrometer  by  comparing  the pyrometer reading with a user-defined target sur- face temperature (termed as the target tempera- ture hereafter). The powder bed was a dried paste, which  was  prepared  by  delivering  dental  powder paste onto a steel substrate and the paste became dried  when  the  substrate  was  heated.  The  steel substrate  could  also  pre-heat  the  powder  bed  to 400 8 C  during  the  laser  densification  process  to reduce or eliminate thermal shock-induced cracks in porcelain bodies. The typical laser experimental conditions  included  the  laser  beam  diameter  of 1 mm, the scanning rate 0.24 mm/s, the chamber pressure  700 Torr,  and  the  user-defined  target temperature 800–1100 8 C with an equivalent laser output power of about 3–20 W.
In  order  to  establish  the  relationship  between the microstructure of laser-densified and furnace- sintered  dental  porcelains,  a  thermogravimetric analyzer  (TGA)  system  was  utilized  to  sinter  the dental porcelain powder. Eight sintering tempera- tures 750, 800, 900, 1100, 1200, 1300, 1400, and 1500 8 C  were  investigated  with  a  heating  rate  of 10 8 C/min,  a  cooling  rate  of  50 8 C/min,  and  a holding time of 5 min. The  as-received  dental  porcelain  powder,  the laser-densified  and  TGA  furnace-sintered  samples were  mounted  using  an  epoxy  resin,  followed  by grinding  and  polishing  down  to  1 m m  diamond suspension.  The  laser-processed  samples  were mounted  and  then  cut  in  such  an  orientation  to reveal the cross-section perpendicular to the laser scanning direction. To determine the existence of the  leucite  phase,  the  polished  samples  were etched  using  1%  hydrofluoric  acid  for  20 s.  An environmental scanning electron microscope (PHI- LIPS ESEM 2020) was used to examine the micro- structure  of  all  the  samples  at  as-polished  and etched  conditions  after  being  carbon  sputter- coated.   The   ESEM   images   with   the   sam magnification  were  analyzed  using  a  Context Vision w multi-mode  image  analyzer  to  determine the area/volume fraction of voids in samples not etched, or the leucite phase in etched samples. For furnace-sintered samples at least 10 different ESEM microphotos were taken to perform the quantitat- ive  image  analysis  and  the  mean  value  was reported.  For  laser-densified  dental  bodies,  12 different locations were selected, and at least five ESEM micrographs were taken around each location in order to perform the quantitative image analysis. Through this procedure, the volume fraction of the leucite  phase  in  laser-densified  samples  as  a function of the location with respect to the center of the scanning laser beam was determined. To  evaluate  the  phase  transformation  as  a function of the TGA holding temperature and relate the microstructural evolution to the phase trans- formation, the TGA furnace-sintered dental porce- lain  chunks  were  pulverized  using  a  mortar  and pestle, and the powders so obtained were analyzed using  X-ray  diffraction  (XRD)  to  investigate  the existence of crystalline phases..