Rate-controlled sintering of ultrafine nickel powder

Rate-controlled sintering of ultrafine nickel powder

NanoShwtured Materials, Pergamon Vol. 5, Nos. 7/8, pp. 835-843.1995 Elsevim Science Ltd Copyright Q 1995 Acta Metallurgica Inc. Printed in the USA...

521KB Sizes 0 Downloads 135 Views

NanoShwtured

Materials,

Pergamon

Vol. 5, Nos. 7/8, pp. 835-843.1995 Elsevim Science Ltd Copyright Q 1995 Acta Metallurgica Inc. Printed in the USA. All tights reserved 0969773l95 $9.50 + .oa

09659773(95)00293-6

RATE-CONTROLLED SINTERING OF ULTRAFINE NICKEL POWDER A.V. Ragulya and V.V. Skorokhod

Institute for Problems of Materials Science, 3, Krzhizhanovsky-Str., 252680, Kiev, Ukraine (Accepted August 1995) Abstract-Results of an experimental investigation on rate-controlled sintering (RCS) of ultrafine nickel powder are considered. Optimization of the temperature-time pathfor sintering allowed a specific density level not less than 0.99 to be achieved, and a grain size less then 100 nm to be maintained (starting grain size was equal 25 nm). An analogous result cannot be attained by constant rate of heating (CRH) schedules. Comparison of the kinetics and structural development of both RCS and CRH processes shows that the RCS method gives much nwre homogeneous and uniform pore shrinkage, without intensive recrystallization, up to a high density level.

INTRODUCTION

Nanocrystalline powders are known to show very rapid densification kinetics in the initial stage of sintering during heating. This phenomenon is accompanied by the formation of large stable pores and local intensive coarsening of the grain structure (1). Such peculiarities demand unconventional approaches to sintering, favorable for enhancing the elimination of pores and restricting grain growth. One contemporary trend in unconventional sintering techniques is connected with rate-controlled sintering (FXS) as a morphology control concept. This method of non-isothermal, non-linear sintering schedule has attracted the attention of ceramists due to its wide possibilities for structural control. Specimens sintered under RCS conditions are characterized by finer grain sizes and pore-free structures. Such results were attained by establishing a feedback between the sintering temperature and the instantaneous density, with the goal of densification rate control at each stage of the sintering process. Authors of the RCS concept proposed the heuristic idea (2): there is a maximum safe rate of shrinkage for each level of fractional density. This restriction of the shrinkage rate is necessary to control both pore and grain structure evolution. Earlier, the main differences in porous structure evolution during sintering under linear heating conditions were analyzed (3). Authors supposed that different micromechanisms of structural evolution (for example, coalescence of pores and their shrinkage, collective recrystallization or grain growth, phase transformations in multicomponent systems, etc.) were occurring 835

836

AV

RAGULYA AND

VV

SKOROKHOD

concurrently, and because of this competition the sir&ring process could be optimized through manipulation. In particular, pore evolution should be looked at carefully throughout sintering, not only for the final stage as in (4) so that a complete experimental data base can be obtained for structure development under non-isothermal conditions. To date, the methodology of rate-controlled sintering has been mainly applied to ceramic materials (5,6), but for metallic powders the number of investigations is negligible (7,8). However, the controlled sintering of carbonyl iron powder (8) displayed great temperature-time mode dependence of the final grain structure and phase distribution. From previous extensive experience, it has become clear that the complete densification of fine powder, especially ultrafine powders of metals, during conventional free sintering is not an ordinary task. It has been shown that linear heating up to high homologous temperatures (T/T,,,&, or isothermal sintering does not result in high density level (residual porosity 6-7%) because of pore coarsening and intensive grain growth. Therefore, this paper compares the sintering of an ultrafine (UF) nickel powder under both rate- controlled conditions and constant heating rate conditions, and analyzes the two processes from both kinetic and structural viewpoints. MATERIALS

AND METHODS

Ultrafine nickel powder produced by a metal evaporation technique (Ultram Ltd., Moscow) was selected as the object of the experiments. To examine the sintering characteristics of the above mentioned powder, the specific surface area measurements (BET), scanning and transmission electron microscopy (SEM andTEM) observations, as well as granulometry were carried out. The impurity contents of oxygen and carbon were analyzed as well. To purify powder from oxygen, low-temperature annealing at temperatures up to 180°C for 1 h in a stream of dry hydrogen was used. Measured properties are summarized in Table 1. The average particle size of Ni powder is equal to 0.025 pm from the data of transmission microscopy analysis. A tendency towards a regular polyhedral shape with smoothed edges can be seen in Figure 1. Residual oxygen concentration after firing in hydrogen was no higher than

TABLE 1 Granulometry and Chemical Composition of UF-Powder Powder

Ni

S,,,

ds,

dIll,

Cont. imp. mass%

m2/g

um

um

[Ol

[Cl

20.58

0.033

0.02

6.25

0.01

0.04

S, - specific surface area, 4, d,, d, - average diameters of particles as determined from specific surface area measurements, microscopy, granulometry data, respectively.

and

SINERING OFULTRAFINE NICKEL POWDER

a37

50 nm Figure 1. TEM micrograph of UF- nickel powder.

Figure 2. Optimized multizoned densification rate-density (D) profiles for RCS and CRS schedules.

0.6-0.8 mass%, but additional purification can be achieved during vacuum sintering. Cylindrical samples with diameter 4.5 mm and height 67 mm were compacted in a steel die at 100-400 MPa. The initial porosity of the green samples was about 67%. A new laboratory of unconventional sintering methods was recently founded in the Institute for Problems of Materials Science (Kiev, Ukraine). Precision dilatometric devices equipped with acomputeraideddesign(CAD) systemandasystem fortherationaloptimizationoftemperature-time paths was elaborated. As developed at IPMS, the precision dilatometer is now utilized for controlled sintering at temperatures up to 2000°C and heating rates up to 2O”C/s. A system for computer analysis of microstructure, grain size and pore size was installed as well.

838

AV

RAGULYA AND

VV

SKOROKHOD

The investigation of sintering kinetics was provided by means of a high temperature vacuum dilatometer (-10e5 Pa). Specimens were heated up to 1100°C with constant rates and subsequently cooled without isothermal exposure. To obtain sufficient data base for rate-controlled firing optimization, the non-isothermal sintering experiments were conducted within a wide range of heating rates: 0.1~6O”C/s. The kinetics field response of ultraline nickel powder was calculated in Ahrrenius coordinates and used for schedule optimization, Microstructure of the green and sintered compacts was characterized by mercury intrusion porometry and scanning electron microscopy. On the basis of mercury porometry (“Autopor9200”, COULTRONICS) the calculation of the pore size distribution functions was made for specimens with several intermediate densities to observe the pore structure development. Microstructure was studiedon broken samples on an REM JSMT-20 (JEOL) andaTEMHU200F (HITACHI) (samples were fractured at 77 K to exclude excess plastic deformation). For the green and sintered samples heated up to several intermediate temperatures, the sizes of the coherent scattering fields were determined by conventional measurement of X-ray line broadening using Cr-Ka radiation. Nickel foil was used as a diffraction standard after heat treatment at 1000°C for 1 h in hydrogen. RESULTS Following the general methodology of RCS-path design, the data base of kinetics responses was formed utilizing preliminary experiments on sintering under linear heating conditions. Such experiments showed that the best final density was developed at heating rate of 0.52 K/s. Two temperature-time paths of controlled sintering were calculated: the first CRS-mode (constant rate sintering) supporting the shrinkage rate about 0.05%/s within fractional density range 40 - 87%, and the second RCS-mode consisting of three stages (should be considered as typical). Figure 2 illustrates the principal differences between these optimized sintering modes. The typical RCS multizonedpath utilizes the fastest rates in the initial stage (at fractional density below OS), slower rates at the intermediate stage (at fractional density between 0.5 and 0.78), and then a regime of continually decreasing rates for the final stage of sintering (fractional density > 0.78). For the CRS schedule, the densification rate remains 1.5-2 times higher than for the RCS case. Nevertheless, one can see that both modes developed high final fractional densities exceeding 0.985. Hydrostatic weighing of the specimens and Hg-porometry analysis shows that the residual porosity is closed. It is interesting to point out that closed porosity appears during controlled sintering at a general porosity level less than 14% and stays negligible (less than 1.3-1.6%) till the end of full densification, whereas under conventional linear heating regime, pore isolation occurs approximately at 17-18% of general porosity and leads to considerable increase closed porosity up to 5-6%. Figures 3 and 4 present pore size distribution for conventional sintering mode (constant rate of heating-CRH) as well as controlled sintering process, respectively. From comparison of these pore size distributions one can conclude that the initial densilication stages are similar up to 650-700°C: the average pore diameter remains invariable and the number of small pores decreases. During CRH process at temperatures higher than 7OO”C, the pore coarsening predominates over pore shrinkage and coarse pores become stable. On the contrary, during the rate-controlled processes a uniform decrease of pore volume occurs over the whole density and temperature ranges and there is a balance between pore growth and shrinkage.

SINTERING OF ULTRAFINE NICKELPOWDER

839

d, pm

Figure 3. Pore size (d) distribution (the percentage of the number of opened pores) in the sintered UF-nickel samples after heating with the rate of 0.44 K/s up to temperature: l- 20, 2- 450,3- 550,4- 600,5- 700,6- 800,7- 900,8- 1000,9- 1100°C.

m40-

3u% zu-

Ii?-

0.i 1 Figure 4. Pore size (d) distribution in the sintered UF-nickel samples after RCS-schedule: l- 20,2- 625,3- 800,4- 930°C.

1:0

840

AV

RAGULYA AND

VV

SKOROKHOD

800 T, =C

4UU-

0

I--

0.2

0.3

i

0.4

I

0.5

I

0.6

I

I

0.7

0.8

I

0.9

I

7.0

D Figure 5(a). Auxiliary temperature-density (D) functions calculation l- RCS, 2- CRS modes.

for

t, m/n Figure 5(b). Temperature-time

paths of l- RCS, and 2- CRS regimes.

Within the temperature interval 480-700°C the pore structureevolution under CRHandCRS modes is similar despite the fact that the densification rate for CRS-mode (0.05%/s) remains less than for CRH-mode at the initial stage only (maximum shrinkage rate is 0.1 %/s at heating rate 0.52 K/s). Thereby, the maximum safe rate should not be exceeded mainly at the initial stage of

841

SINTERING OF ULTRAFINE NICKELPOWDER

80-

0.2

0.4

0.6 D

0.8

1.0

Figure 6. Density (D): dependencies of the coherent scattering field for sintered ultrafine nickel powders after: l- rate-controlled mode, 2- constant rate of heating mode (0.44 K/s).

sintering. The temperature-time schedules of both RCS and CRS runs are very close (Figure 5) insofar as the ratio of the heating rate to densification rate remains constant. In the RCS experiments, though, density levels close to theoretical were reached at temperatures of 10201030°C, which are less than final temperature in the CRH preliminary experiments (1100°C). The achievement of full density at lower temperatures is favorable for preventing grain growth. As evidence of the lack of grain growth, density levels close to theoretical were reached at temperatures of 1020-1030°C which is less than the final temperature in the CRH preliminary experiments (1 100°C). The constancy of the average pore size throughout RCS densification can be evidence of the lack of grain growth. The electron microscopy data and sizes of the coherent scattering field also confirm the smaller grain sizes for the RCS-sintered samples. If one assumes the ultrafine particles are monocrystals and that the remnant elastic stresses are relaxed at low temperatures, one can conclude that the size of coherent scattering field is equal the grain size, and that the RCS method produces finer grain sizes for a given density (Figure 6). The microstructure of the broken sample after heating up to 900°C under RCS conditions is shown in Figure 7. For the linear heating schedule, grain growth starts above 700°C but for the RCS-mode, grain boundary migration is suppressed by the network of open pores up to 800-820°C. Density dependencies of grain size throughout RCS and CRH sintering are compared in Figure 6 and Table 2. There is a constancy in the relation between average pore size and average grain size within the whole density range for RCS mode and within the restricted range (up to 700-800°C) for linear heating.

AV RAGULYA

842

AND

VV SKOROKHOD

Figure 7. Scanning electron micrograph of sintered UF-nickel powders under rate-controlled conditions (930°C).

TABLE 2 Geometry Characteristics of Porous Structure after RCS- and CRH-Sintering. Constant rate of heating

T

D

T 20 4.50 550 600 700 800 900 1000

0.33 0.37 0.49 0.55 0.65 0.73 0.82 0.86

Rate-controlled sintering

I

R,

d,

d/R

Pm

w

Pm

0.025 0.027 0.028 0.028 0.030 0.033 0.042 0.067

0.040 0.042 0.044 0.045 0.047 0.052 0.112 0.225

1.60 1.55 1.56 1.61 1.57 1.58 2.33 3.32

T, I

D

“C

R,

d,

w

Pm

d/R

20 450

0.33 0.37

0.025 0.027

0.040 0.042

1.60 1.55

625

0.63

0.029

0.046

1.55

800 930 1030

0.81 0.91 0.99

0.031 0.040 0.076

0.049 0.062 -

1.58 1.55 -

D- density, R- average grain size, d- average pore size.

The magnitude of the d/R relation is approximately equal to 1.55 and is in contradiction with the known results for coarse grain structure and porosity level less then 0.4, where d/R< 1. The initial porosity of the investigated green samples was 0.67, so the average pore diameter was greater than average grain diameter. After the RCS process, the grain size remains less than 0.1 pm but for the CRH-mode, grain growth yields a grain size considerably greater than 0.1 pm, with the result that further den&cation is restrained. Under a rate-controlled sintering regime the processes of pore coalescence and grain growth are suppressed and densification occurs by coherent sintering. Thus, the optimal temperature-time regime of sintering reflects an optimal path of structural development for both pores and grains.

SINTERING OF ULTRAFINE NICKELPOWDER

843

SUMMARY

When sintered under rate-controlled conditions, the ultrafine nickel powder has been densified to an almost pore-free state with a grain size not exceeding 100 nm, as can be seen by the constancy of the ratio between pore and grain sizes, which illustrates the near-coherent character of the rate-controlled sintering process. Further development of rate-controlled sintering may prove advantageous for the creation of dense materials with nanocrystalline structure.

REFERENCES 1. 2. 3. 4. 5. 6. 7. 8.

D.-J. Chen and M.J. Mayo, Nanostruct. Mater. 2,469 (1993). H. Palmour III, D.R. Johnson, Sintering and Related Phenomena, Gordon & Breach Publishers, New-York, p.779 (1967). A.V. Ragulya, Kinetics and Mechanisms of Sintering of Ultrajine Powders with Different type of Chemical Bonding under High rates of Heating, Ph.D. Thesis, Kiev, 182 pages (1992). H. Palmour III and T.M. Hare, Proc. of the Sixth World Round Table Conference on Sintering, Hercig-Novi (1985), New-York, Plenum Press, p. 16 (1987). H. Palrnour III and M.L. Huckabee, Materials Science Research, NewYork, Plenum Press, Vol. 6, p. 257 (1973). H. Palmour III, M.L. Huckabee andT.M. Hare, Proc. of the Fourth WorldRound Table Conference on Sintering, Material Science Monographs, Amsterdam, Elsevier, Vol. 4, p. 46 (1979). H. Palmour III, Powder Metal Report, N9, p. 572 (1988). H. Palmour III, Proc. of the Seventh World Round Table Conference on Sintering, Hercig-Novi 1989, New-York, p. 337 (1989).