Improvement of the Dimensional Stability of Powder Injection
Molded Compacts by Adding Swelling Inhibitor
into the Debinding Solvent
YANG-LIANG FAN, KUEN-SHYANG HWANG, and SHAO-CHIN SU
Defects are frequently found in powder injection molded (PIM) compacts during solvent debinding due to the swelling of the binders. This problem can be alleviated by adjusting the composition of the debinding solvent. In this study, 10 vol pct swelling inhibitors were added into heptane, and the in-situ amounts of swelling and sagging of the specimen in the solvent were recorded using a noncontacting laser dilatometer. The results show that the addition of ethanol, 2-propanol, 1-butanol, and 1-pentanol reduced the amounts of swelling by 31, 21, 17, and 11 pct, respectively. This was because the small molecule alcohols, which do not dissolve par-affin wax (PW) or stearic acid (SA) in the binder system, could diffuse easily into the specimen and increased the portion of the swelling inhibitor inside. The amount of the extracted PW and SA also decreased, but only by 8.3, 6.1, 4.3, and 2.4 pct, respectively. The solubility parameters of 1-bromopropane (n-PB) and ethyl acetate (EA) are between those of heptane and alcohols, and they also yielded a slight reduction in the amounts of swelling by 6 and 11 pct, respectively. These results suggest that to reduce defects caused by binder swelling during solvent debinding, alcohols with high solubility parameters can be added into heptane without sacrificing signifi-cantly on the debinding rate.
DOI: 10.1007/s11661-007-9351-y
The Minerals, Metals & Materials Society and ASM International 2007
I. INTRODUCTION
THE
two-stage debinding method, solvent debinding followed by thermal debinding, is one of the most widely adopted processes in the current powder injection molding (PIM) industry.[1]During the solvent debinding process, soluble binder components such as paraffin wax (PW) and stearic acid (SA) are extracted by the solvent and leave interconnected pores. This facilitates the removal of the decomposed gas molecules when non-soluble backbone binders, such as polyethylene (PE), are subjected to the subsequent thermal debinding.[2,3] However, some defects and poor dimensional control are frequently encountered in the solvent debinding stage. For example, PIM parts with areas of different thickness will have different amounts of expansion in each of those different areas. This is because polymeric binders swell in the solvent and the amount of swelling depends mainly on three factors: thickness of the specimen, content and type of the binder, and solvent temperature.[4] Parts without flat supports or with cantilever sections may even slump or distort during debinding, when the solvent diffuses into the binder and causes swelling and softening of the part.[4,5] Such polymer swelling can be explained by thermodynamicparameters, particularly the change in free energy of the polymer solution.[6]
DG¼ DH TDS ½1
where DG, DH, and DS are the changes in the free energy, enthalpy, and entropy of mixing, and T is the absolute temperature. When a polymer is immersed in the solvent, entangled polymers are separated into many individual polymer chains and therefore the entropy of mixing of the polymer solution is usually positive (DS > 0). However, each segment on the single polymer chain is still linked by the covalent bond. Thus, the entropy of mixing for polymer solutions remains small. This suggests that the free energy of mixing in Eq. [1] is mainly dependent on the enthalpy change.
Hildebrand derived an equation to relate the enthalpy change, DH, with the solubility parameter d as[7]
DH¼ V/s/pðdp dsÞ2 ½2 where V is the volume of the solution; /sand /pare the volume fractions of solvent and polymer, respectively; and ds and dp are the solubility parameters of solvent and polymer, respectively. The solubility parameter describes the affinity between molecules of the materials, and it can be quantified using the following equation:[6]
d¼ DðDHv RTÞ M
1=2
½3
where DHV, M, D, R, and T are heat of evaporation, molecule weight, density, gas constant, and absolute
YANG-LIANG FAN and SHAO-CHIN SU, Graduate Students, and KUEN-SHYANG HWANG, Professor The Department of Materials Science and Engineering, National Taiwan University, Taipei, 106, Taiwan, Republic of China. Contact e-mail: kshwang@ccms.ntu.edu.tw
Manuscript submitted March 31, 2007. Article published online December 28, 2007
temperature, respectively. This equation, however, can-not be used to determine the solubility parameters for cross-linked polymers because they usually decompose before reaching the boiling point and measurement of the heat of evaporation. Thus, the swelling technique is often used for the determination of solubility parameters of polymers.[8]
According to Eq. [2], when the solubility parameter of the polymer is close to that of the solvent, the DH becomes small and gives a negative DG, which enhances the polymer dissolution. In general, if dp ds
< 1.1 is satisfied, the polymer is more likely to be soluble in the solvent unless its molecular weight is too large.[6]In such a case, solvent molecules can penetrate easily into the holes and cavities in the entangled polymers, and induce polymer swelling and even dissolution. Using the binder components examined in this study, the solubility parameters of PW and PE are 7.8 and 7.9 H, respec-tively.[9,10] Thus, the PW can be dissolved easily in heptane, which has a solubility of 7.4 H, while the PE is not soluble.
Alcohol molecules have a hydroxyl functional group (-OH) and tend to form hydrogen bonding between two molecules. Thus, the solubility parameter of alcohol is always higher than those of nonpolar solvents. How-ever, this parameter decreases when the number of carbon increases. For instance, the solubility parameter of ethanol is 12.7 H, but that of 1-pentanol is 11.0 H.[11] Considering the large difference in the solubility eters of alcohol and PE, which has a solubility param-eter of 7.9 H, the swelling of PIM parts in heptane should be suppressed when alcohol is added. However, the debinding rate will also decrease, since alcohol does not dissolve PW. Therefore, in order to use alcohol as a swelling inhibitor, it is necessary to optimize the type and the amount added. In this study, four different alcohols, ethanol, 2-propanol (IPA), 1-butanol, and 1-pentanol, were added into heptane as swelling inhib-itors. In addition, 1-bromopropane (n-PB) and ethyl acetate (EA), which dissolve PW, were also added as swelling inhibitors. Their effects on debinding behavior were examined through the in-situ length change, in-situ sagging distance, and amount of binder removal. These examinations provide a solid reference in selecting a good swelling inhibitor that could mitigate the amount of swelling during solvent debinding while maintaining the debinding efficiency.
II. EXPERIMENTAL PROCEDURE
A carbonyl iron powder (CIP-S 1641, ISP Corp., Wayne, NJ) with an average particle size of 4.88 lm and a pycnometer density of 7.57 g/cm3was used as the base powder in this study. To prepare the feedstock, 93 wt pct of metal powders were kneaded with 7 wt pct of multicomponent binders, which consisted of 40 wt pct PE (USI Far East Corp., Taipei, Taiwan), 55 wt pct PW (Nippon Seiro Corp., Tokyo), and 5 wt pct SA (Nacalai Tesque Inc., Kyoto, Japan). To find the melting points of these binder components, which are good references for determining the molding and solvent debinding temperatures, a differential scanning calorimeter (DSC 910, Du Pont Corp., Wilmington, DE) was used. The measurement was carried out under nitrogen atmo-sphere, and the heating rate was 5 C/min. The resulting feedstock was then molded at about 140C into rectangular specimens 4· 10 · 100 mm in size using an injection molding machine (270C, Arburg GmbH, Lossburg, Germany). To prepare the debinding solvent, 10 vol pct swelling inhibitors, methanol, ethanol, IPA, 1-butanol, or 1-pentanol, were mixed with heptane for 5 minutes. However, since separation was noticed between methanol and heptane, only ethanol, IPA, 1-butanol, and 1-pentanol were examined. In addition to alcohols, n-PB and EA, which also dissolve PW, were also evaluated as swelling inhibitors. The properties of the heptane, swelling inhibitors, and binder components are given in TableI. Solvents in this study were purchased from Aldrich Corp. (St. Louis, MO) and were used without any purification.
To understand the effects of swelling inhibitors on debinding, the in-situ dimensional changes of speci-mens were measured using a self-designed laser dila-tometer. The setup and the details of the instrument can be found in previous articles.[4,5] The temperature of the laser dilatometer was controlled at 40C ± 0.2 C. For measurement of the sagging distance, the specimen was supported by two cylindri-cal blocks, which were separated by a distance of 80 mm.[4] At least two specimens were examined in each solvent, and the results were consistent. Thus, only the first set of data was reported. For measure-ment of the amount of binder removed during solvent debinding, at least four specimens were measured and their averages reported.
Table I. Properties of Solvents and Binder Components Used in This Study Material Solubility Parameter (cal/cc)1/2[9–11,18] Density (g/cm3)[11] Boiling Point (C)[11]
Molar Volume (cm3/mol)
Heptane 7.4 0.68 98.1 147.4 1-pentanol 11.0 0.80 140.0 110.5 1-butanol 11.4 0.80 117.7 92.5 IPA 11.9 0.79 97.2 76.8 Ethanol 12.7 0.78 78.3 58.5 n-PB 8.9 1.35 71.1 91.1 EA 9.1 0.91 77.1 96.5 PE 7.9 0.92 — >10,000 PW 7.8 0.91 — 433.0 SA 7.5 0.96 — 295.8
III. RESULTS
A. Swelling Behavior of Pure PE
Figure 1 shows that the melting points of the PE and PW, which were measured using a DSC, were about 93 and 58C, respectively. The peak at 42 C was the solid-solid phase transition of PW.[12] Our previous results have shown that when the debinding temperature is higher than the melting point of the binder components, the specimen will exhibit a large amount of swelling.[4] Therefore, the debinding temperature selected in this study was set at 40 C.
Figure 2 shows the swelling curves of pure PE in different solvents. When the specimen was immersed in heptane, the amount of swelling increased as immersion time increased. In pure alcohols, however, the amount of swelling quickly reached equilibrium at about 0.3 pct,
which was attributed to thermal expansion. This sug-gested that PE does not have a swelling reaction with alcohols. This confirms the results reported by Makitra et al., who showed that the amount of swelling of PE in heptane is about 1,400 times that in ethanol.[13] Since alcohols do not cause swelling in PE, their additions of 10 vol pct to heptane inhibited the amount of swelling of PE slightly, as shown in Figure 2.
B. The Swelling Behavior of PIM Specimens in Heptane + Alcohol
Figure 3 shows the swelling behavior of molded parts in different solvents. The specimen swelled in the beginning due to thermal expansion.[4] As solvent molecules started to diffuse into the specimen, soluble binder components (PW and SA) began to dissolve. At this stage, the dissolved binders were close to the surface and thus could diffuse out easily,[14] leading to little swelling. As the debinding time increased to about 8 ks, the amount of swelling increased quickly and then reached a plateau at about 12 ks. The maximum amount of swelling of the specimen in pure heptane was about 0.84 pct. When 10 vol pct ethanol, IPA, 1-butanol, and 1-pentanol were added, the amounts of swelling decreased to 0.58, 0.66, 0.70, and 0.75 pct, respectively, or a reduction of 31, 21, 17, and 11 pct, respectively. These results showed that alcohols acted as swelling inhibitors in the solvent debinding process and that their effectiveness decreased with increasing numbers of carbons.
Figure 4 shows that the sagging distance of the specimen also decreased when alcohols were added. The amounts of sagging were 0.439, 0.456, 0.531, and 0.585 mm, respectively, for the addition of 10 vol pct ethanol, IPA, 1-butanol, and 1-pentanol. Compared to the sagging distance in pure heptane, these distances were equivalent to reductions of 29, 26, 14, and 5 pct, respectively. The trends of these curves were similar to those of the swelling curves shown in Figure 3.
40 -0.20 -0.15 -0.10 -0.05 0.00 PW-Solid Phase Transition PW melting peak PE melting peak 42°C 58°C Temperature (°C) ) g/ w( w ol F t a e H 93°C 120 100 80 60
Fig. 1—DSC curve of the molded part.
0 0 1 2 3 4 Time (ks) ) %( e g n a h C ht g n e L heptane heptane-10% 1-pentanol heptane-10% 1-butanol heptane-10% IPA heptane-10% ethanol 1-pentanol 1-butanol IPA ethanol 25 20 15 10 5
Fig. 2—Amount of expansion of pure PE specimen in different sol-vents at 40C. 0 -1.0 -0.5 0.0 0.5 1.0 Time (ks) ) %( e g n a h C ht g n e L heptane heptane-10% 1-pentanol heptane-10% 1-butanol heptane-10% IPA heptane-10% ethanol 15 10 5
Fig. 3—Amount of expansion decreases as 10 vol pct of 1-pentanol, 1-butanol, IPA, and ethanol is added into heptane.
C. Effect of Alcohol Addition on Debinding Completion
The solubility of PW in the low polarity heptane at 40C was high, at about 1040 g/L. However, in high polarity alcohols, it is usually low (about zero in ethanol and about 12 g/L in 1-pentanol). Therefore, it was expected that adding alcohols into heptane would decrease the debinding rate. Figure 5 shows that after 18 ks, the percentages of soluble binders (PW and SA) that were removed reached 66.2, 68.4, 70.2, and 72.1 pct, respectively, when 10 vol pct ethanol, IPA, 1-butanol, and 1-pentanol were added. Compared to that in heptane, the binder removal rate was reduced by 8.3, 6.1, 4.3, and 2.4 pct, respectively. Despite these reductions, the amounts of binder removal were still
adequate since the generally accepted minimum amount after solvent debinding is 60 to 70 pct.[15]
Based on the changes in specimen length, sagging distance, and amount of binder removed, the ethanol addition had the best inhibition effect in swelling and sagging (about 30 pct) but its debinding efficiency was the lowest, reduced by 8.3 pct compared to that in pure heptane.
D. Swelling Behavior of PIM Specimens in Heptane + n-PB or EA
In addition to the evaluation of alcohol additives, which do not dissolve PW or SA, the addition of n-PB and EA, which are good solvents for PW and SA, were also examined. The measured solubilities of PW in 40C n-PB and EA were about 1120 and 320 g/L, respec-tively. Figure 6 shows that the maximum amounts of swelling after adding 10 vol pct n-PB and EA were 0.79 and 0.75 pct, respectively, or a reduction of 6 and 11 pct compared to that in heptane. Since the molecular size of 1-butanol is close to that of n-PB and EA, the swelling curve in heptane + 10 vol pct 1-butanol is also shown in Figure 6 for comparison. Similar to that of the length change, the maximum amount of sagging distance (Figure 7) decreased only slightly when 10 vol pct n-PB or EA was added into heptane.
Figure 8 shows the percentages of the soluble binders (PW and SA) that were removed when 10 vol pct n-PB, EA, and 1-butanol was added into heptane. Since n-PB has a higher solubility for PW than heptane, the debinding completion increased slightly in the hep-tane + 10 vol pct n-PB. For example, after 18 ks, the debinding completion increased from 74.5 to 76.3 pct. When 10 vol pct EA, which has a lower solubility for PW than heptane, was added, the debinding completion decreased from 74.5 to 73.9 pct, as expected. The addition of 10 vol pct 1-butanol decreased more notice-ably from 74.5 to 70.2 pct. 0 0.0 0.2 0.4 0.6 Time (ks) ) m m( e c n at si D g ni g g a S heptane heptane-10% 1-pentanol heptane-10% 1-butanol heptane-10% IPA heptane-10% ethanol 15 10 5
Fig. 4—In-situ sagging distance of specimens decreases after adding 10 vol pct swelling inhibitor.
0 0 20 40 60 80 100 Debinding Time (ks) r e d ni B el b ul o S f o e g at n e cr e P ) %( d e v o m e R heptane heptane-10% 1-pentanol heptane-10% 1-butanol heptane-10% IPA heptane-10% ethanol 80 60 40 20
Fig. 5—Effect of adding 10 vol pct ethanol, IPA, 1-butanol, and 1-pentanol on the amount of soluble binder removed during solvent debinding in heptane 40C. 0 -0.5 0.0 0.5 1.0 Time (ks) ) %( e g n a h C ht g n e L heptane heptane-10% nPB heptane-10% EA heptane-10% 1-butanol 20 15 10 5 5
Fig. 6—Amount of expansion decreases slightly after adding 10 vol pct n-PB and EA into heptane.
E. Inhibition of Swelling Using Low Debinding Temperatures
Although the amount of swelling can be reduced by adding swelling inhibitors into heptane, the same effect can also be achieved by lowering the debinding temper-ature.[4] Therefore, it would be worthwhile to check which method is better by comparing the amount of swelling and the amount of binder removed. Figure 9 shows that the amount of swelling increased as the debinding temperature increased. Also shown in Figure 9 are the maximum amounts of swelling mea-sured in 40C heptane that contained 10 vol pct ethanol and IPA, and these values were equivalent to those attained in pure heptane at 32 and 34C, respectively.
Figure 10 compares the effect of adding swelling inhibitors in heptane and using different debinding temperatures. Using 60 pct binder removal as an example, which is generally accepted as the minimum amount
required before moving on to the subsequent thermal debinding step, 12.0 ks was required to reach this amount when 40C pure heptane was used. When 10 vol pct ethanol and IPA were added into heptane, the debinding time required to reach the same 60 pct increased slightly to 14.1 and 13.6 ks, respectively. In contrast, with 32C and 34 C heptane, the debinding time required to attain the same 60 pct binder removal increased significantly to 52.5 and 44.7 ks, respectively. These results indicated that to reduce the amount of swelling during solvent debinding, the addition of swelling inhibitors was a more effective method than lowering the debinding temperature.
IV. DISCUSSION
A. Solubility Parameter of Molded Part
The binder system that was used in this study consisted of PE, PW, and SA. Polyethylene, which has 0 0.0 0.2 0.4 0.6 Time (ks) ) m m( e c n at si D g ni g g a S heptane heptane-10% nPB heptane-10% EA heptane-10% 1-butanol 15 10 5
Fig. 7—In-situ sagging distance of specimens decreases after adding swelling inhibitors, 10 vol pct n-PB and EA, into heptane.
0 20 40 60 80 0 20 40 60 80 100 Debinding Time (ks)
Percentage of Soluble Binder
Removed (%)
heptane
heptane-10% nPB heptane-10% EA heptane-10% 1-butanol
Fig. 8—Effect of adding 10 vol pct n-PB and EA on the amount of soluble binder removed during solvent debinding at 40C.
30 0.5
1.0
32
Solvent Debinding Temp. (°C)
f o t n u o m A . x a M ) %( g nill e w S heptane heptane-10% IPA heptane-10% ethanol 34 50 40
Fig. 9—Effect of solvent debinding temperature and addition of swelling inhibitors on amount of swelling of PIM specimens.
0 0 20 40 60 80 100 Debinding Time (ks) r e d ni B el b ul o S f o e g at n e cr e P ) %( d e v o m e R heptane (40°C) heptane (34°C) heptane (32°C) heptane-10% IPA (40°C) heptane-10% ethanol (40°C) 100 80 60 40 20
Fig. 10—Effect of debinding temperature and addition of swelling inhibitors on percentages of soluble binders extracted.
a solubility parameter of about 7.9 H,[9] is frequently used as one of the backbone binders in PIM parts. Since the difference in solubility parameters between heptane and PE is small dPE dheptane
¼ 0.5 < 1.1, the attraction forces between PE chain segments are smaller than those of the polymer-solvent interactions. More-over, polymer molecules usually consist of long chains or coils with large numbers of segments that are tightly folded or entangled with each other. Thus, there are tortuous channels and small cavities inside the polymer specimen.[16] As a result, when PE is immersed in heptane, it absorbs solvent molecules, thus causing the polymer matrix to swell, as shown in Figure 2. However, as ethanol or IPA is added into heptane, the small channels and cavities are selectively filled with these small molecule alcohols. Since the actual content of ethanol or IPA inside the specimen is higher than 10 vol pct, the extent of swelling reduction of a pure PE specimen is thus greater than 10 pct, as shown in Figure 2.
The other major binder component in the PIM specimen is PW. It is composed of saturated hydrocar-bons with ~20 to 30 carbons, and the most popular constituent is C28H58.[17] The solubility parameter of PW is about 7.8 H.[10] Since the difference in solubil-ity parameters between PW and alcohols, such as 1-pentanol, dPW dpentanol
¼ 3.2>1.1is much greater than that between PW and heptane dPW dheptane
¼
0.4<1.1Þ, it is expected that a larger amount of swelling of PW will be found in pure heptane than that found in 1-pentanol-added heptane. The last binder component in the binder system used in this study was SA (C18H36O2), which has a solubility parameter of about 7.5 H.[18]Since its solubility parameter is close to the 7.4 H of the heptane, the amount of swelling should also be significant. With these known solubility parameters and the binder compositions, the solubility parameter of the molded part can be estimated using the mixing rule. The resulting value is about 7.8 H, which is close to that of heptane
dspecimen dheptane
¼ 0.4<1.1. Thus, the amount of swelling of the molded part in the heptane is significant.
B. Using Alcohols as Swelling Inhibitors
Figure 3 illustrates that when alcohols were added into heptane, the maximum amounts of swelling decreased. This indicated that alcohols act as swelling inhibitors and that their effectiveness decreased with increasing numbers of carbons. To aid in discussion of
the effects of alcohols on swelling, the length changes and the sagging distances are summarized in Table II. The data show that as 10 vol pct ethanol was added into heptane, the amount of swelling was reduced by 31 pct, and the amount of sagging distance by 29 pct. As the molecular size of the swelling inhibitor increased, the reduction in swelling decreased. This trend is most likely related to the molar volume and the solubility parameters of the alcohols. Table I shows that as the number of carbons increases, the molar volume of alcohol increases and the solubility param-eter decreases. Among the four alcohols examined, ethanol has the smallest molar volume, the highest solubility parameter, and the best swelling inhibition effect. This was mainly because the solubility parameter of the low molecule alcohol is much larger than those of PE, PW, and SA. In addition, its diffusion rate into the specimen is faster than that of heptane molecules. Thus, the local spaces among PE chains are filled with swelling inhibitors. With a lesser amount of swelling and with more alcohols filling in the cavities inside the backbone binder, PE, the polymer chains become stronger than in pure heptane. Therefore, the recorded amount of sagging distance decreased. These results suggest that ethanol is the best choice among the four alcohols examined.
The lesser amount of swelling, however, causes a slower debinding rate, as shown in Figure 5. One reason is that PW does not dissolve in the 40C alcohols. The other reason is that the openings for the solvent penetration and for the outward diffusion of the PW and SA-containing solution become smaller due to the lesser amount of swelling.
C. Using n-PB and EA as Swelling Inhibitors
Since n-PB and EA dissolve PW and SA, their additions to heptane do not change the binder removal rate much during solvent debinding, as shown in Figure 8. These two solvents also help reduce the swelling and sagging, as shown in Figures 6 and 7, and Table II. However, they are not as effective as alcohols. This suggests that the solubility parameters and molar volume are still the most important proper-ties of the inhibitors. Solvents with high solubility parameters and low molar volumes, such as ethanol and IPA, can inhibit the swelling and sagging more effec-tively than n-PB and EA, which have lower solubility parameters and large molar volumes.
Table II. Effect of Adding Swelling Inhibitors on Amount of Swelling and Sagging of PIM Specimens Debinding Solvent Maximum Amount of Swelling (Pct) Swelling Reduction (Pct) Max. Amount of Sagging (mm) Sagging Reduction (Pct) Heptane 0.84 — 0.614 — +10 pct 1-pentanol 0.75 11 0.585 5 +10 pct 1-butanol 0.70 17 0.531 14 +10 pct IPA 0.66 21 0.456 26 +10 pct ethanol 0.58 31 0.439 29 +10 pct n-PB 0.79 6 0.594 3 +10 pct EA 0.75 11 0.568 8
In summary, the selection of a suitable swelling inhibitor will depend on whether the debinding rate or the amount of reduction in swelling is more important. When debinding completion is the major concern, n-PB should be employed. In contrast, when the reduction in the amount of swelling and sagging is more critical, ethanol should be used.
D. Shrinkage during Solvent Draining
Other than the swelling that can cause dimensional control problems of PIM parts, anisotropic or sudden shrinkage can also cause problems. For example, at the end of the debinding stage, as the debinding solvent was drained out of the specimen, the specimen temperature decreased sharply due to the evaporation of the sol-vent.[5]This induced a large amount of shrinkage in the specimen, as shown in Figures 3 and 4. When solvent inside the specimen evaporated quickly (particularly ethanol, which has a low boiling point and high heat of vaporization) the shrinkage became more apparent. This large amount of shrinkage could induce stresses inside the specimen, causing fish scale–like delaminations and even breakages. In order to alleviate this shrinkage, the debound part should be kept in the solvent tank and the solvent cooled slowly to room temperature before the parts are removed.
V. CONCLUSIONS
The swelling of PIM parts that frequently occurs during solvent debinding usually causes defects, such as cracking and distortion, and problems in dimensional control. These problems can be alleviated by adding alcohols or solvents with high solubility parameters. This study shows that when 10 vol pct ethanol, IPA, 1-butanol, and 1-pentanol are added into heptane, the maximum amounts of swelling decrease from 0.84 of pure heptane to 0.58, 0.66, 0.70, and 0.75 pct, respec-tively. Similar trends are also found in the sagging distance. However, the amounts of binder removed decreased slightly by 8.3, 6.1, 4.3, and 2.4 pct, respec-tively. Adding n-PB and EA into heptane can also inhibit the swelling of the specimen, but this approach is less effective than adding alcohols. These results show
that good swelling inhibitors should possess high solu-bility parameters and small molar volumes. Low tem-perature debinding can also mitigate the swelling. However, the debinding time will be more than tripled when the same amount of swelling and binder removal reached by adding alcohols need to be attained.
ACKNOWLEDGMENT
The authors thank the National Science Council for their support of this project (Contract No. NSC 94-2216-E-002-004).
REFERENCES
1. K.P. Johnson: U.S. Patent 4,765,950, 1998.
2. E.J. Wsetcot, C. Binet, and R.M. German: Powder Metall., 2003, vol. 46, pp. 61–67.
3. K.S. Hwang and Y.M. Hsieh: Metall. Mater. Trans. A, 1996, vol. 27A, pp. 245–53.
4. S.C. Hu and K.S. Hwang: Metall. Mater. Trans. A, 2000, vol. 31A, pp. 1473–78.
5. H.K. Lin and K.S. Hwang: Acta Mater., 1998, vol. 46, pp. 4303– 09.
6. S.F. Sun: Physical Chemistry of Macromolecules: Basic Princi-ples and Issues, 2nd ed., Wiley-Interscience, New York, NY, 2004, pp. 67–79.
7. J.H. Hildebrand and R.L. Scott: Solubility of Nonelectrolytes, Reinhold, New York, NY, 1950, pp. 135–54.
8. M. Sen and O. Gu¨ven: J. Polym. Sci., Polym. Phys., 1998, vol. 36, pp. 213–19.
9. J. Brandrup and E.H. Immergut: Polymer Handbook, 3rd ed., Wiley-Interscience, New York, NY, 1966, pp. VI 347–VI 351. 10. R.B. Richards: Trans. Faraday Soc., 1946, vol. 42, pp. 20–28. 11. R.C. Weast: Handbook of Chemistry and Physics, 86th ed., CRC
Press, Inc., Boca Raton, FL, 2005–2006, chap. 3, pp. 4–522. 12. S.P. Srivastava, J. Handoo, K.M. Agrawal, and G.C. Joshi:
J. Phys. Chem. Solids, 1993, vol. 54, pp. 639–70.
13. R. Makitra, Y. Pyrih, E. Sagladko, A. Turoskiv, and G. Zaikov: J. Appl. Polym. Sci., 2001, vol. 81, pp. 3133–40.
14. T.S. Shivashankar and R.M. German: J. Am. Ceram. Soc., 1999, vol. 82, pp. 1146–52.
15. Y. Li, F. Jiang, L. Zhao, and B. Huang: Mater. Sci. Eng. A, 2003, vol. 362, pp. 292–99.
16. D.J. Williams: Polymer Science and Engineering, Prentice-Hall, Inc, Englewood Cliffs, NJ, 1971, pp. 163–92.
17. F. Bordet, T. Chartier, and J.-F. Baumard: J. Euro. Ceram. Soc., 2002, vol. 22, pp. 1067–72.
![Table I. Properties of Solvents and Binder Components Used in This Study Material Solubility Parameter (cal/cc) 1/2[9–11,18]](https://thumb-ap.123doks.com/thumbv2/9libinfo/8775348.213731/2.890.42.794.876.1064/table-properties-solvents-binder-components-material-solubility-parameter.webp)
