35

Comparison of Erosion Rates of SUS304 and SUS316 Stainless

Steels by Molten Sn–3Ag–0.5Cu SolderIkuo Shohji*, Kazuhito Sumiyoshi* and Makoto Miyazaki**

*Department of Mechanical System Engineering, Graduate School of Engineering, Gunma University, 1-5-1 Tenjin-cho, Kiryu 376-8515, Japan

**Nagano Oki Electric Co., Ltd., 965 Mimitori, Komoro 384-0084, Japan

(Received July 3, 2009; accepted October 20, 2009)

Abstract

An erosion test was conducted to compare the erosion rates of SUS304 and SUS316 stainless steels by molten Sn–3Ag–

0.5Cu (mass%) lead-free solder. Stainless steels attached with polyvinyl chloride were used to accelerate the occurrence

of erosion by destruction of the passivity of stainless steel and to shorten the incubation period of the occurrence of

erosion. The average erosion rate of SUS304 steel is approximately 20% faster than that of SUS316 steel, whereas the

maximum erosion depth evaluated by extreme value analysis does not depend on steel type. From microstructural

observation of erosion interfaces, it was found that Fe–Cr–Sn and Fe–Cr–Mo–Sn layers form at the erosion interfaces of

SUS304 and SUS316 steels, respectively.

Keywords: Stainless Steel, Erosion Rate, Sn–3mass%Ag–0.5mass%Cu, Extreme Value Analysis, Microstructure

1. Introduction Although lead-free soldering has spread to many elec-

tronic instruments, there are still several problems which

have to be settled. One of these problems is the erosion of

stainless steel by molten lead-free solder in flow solder-

ing.[1] Although the mechanism of this erosion has not yet

been clarified, the erosion process seems to consist of two

processes; the destruction of the passivity of stainless steel

and dissolution of stainless steel into the molten solder.

Destruction of passivity strongly depends on various con-

ditions such as the chemical composition and microstruc-

ture of stainless steel, properties of passivity, and so on.

Since the incubation period of the occurrence of erosion

changes depending on such conditions, its evaluation

becomes difficult.[2]

We have developed a new method using polyvinyl chlo-

ride to accelerate the occurrence of stainless steel erosion

by the molten lead-free solder.[3] Since polyvinyl chloride

has a strong flux function, stainless steel attached with

polyvinyl chloride is easily attacked by the molten lead-free

solder. Moreover, the erosion behavior of polyvinyl chlo-

ride on stainless steel was confirmed to be similar to that

of the flux used for soldering. Thus, using polyvinyl chlo-

ride can shorten the incubation period until the occurrence

of erosion of stainless steel.

In this study, an erosion test was conducted to compare

the erosion rates of SUS304 and SUS316 stainless steels by

molten Sn–3Ag–0.5Cu (mass%) lead-free solder. In addi-

tion, microstructural observation of the erosion interfaces

was conducted using an electron probe X-ray microana-

lyzer (EPMA).

2. ExperimentalSUS304 steel samples 85 × 8 × 2 mm and SUS316 steel

samples 105 × 8 × 2 mm were prepared for the erosion test.

Table 1 shows the chemical compositions of the stainless

steels used in this study. Polyvinyl chloride powder

(Wako, n: about 1100) was prepared. The average diame-

Table 1 Chemical compositions of SUS304 and SUS316 stainless steels used.

C Si Mn P S Ni Cr Mo Fe (mass%)

SUS304 0.06 0.44 0.82 0.028 0.006 8.02 18.17 — Bal.

SUS316 0.05 0.70 0.98 0.031 0.006 10.14 16.84 2.02 Bal.

Shohji et al.: Comparison of Erosion Rates of SUS304 and SUS316 Stainless Steels (1/5)

36

Transactions of The Japan Institute of Electronics Packaging Vol. 2, No. 1, 2009

ter of the powder particles was approximately 70 μm. Poly-

vinyl chloride was applied to the surfaces of stainless steels

as follows. First, the specimen of stainless steel was heated

up to 120°C on the hot plate and held for 4 min. The heated

specimen was put into polyvinyl chloride powder and

cooled. After cooling, any excess powder on the surface of

the specimen was removed. Approximately 1.6 mg and 2

mg of polyvinyl chloride were attached to the SUS304 and

SUS316 specimens, respectively.

Figure 1 shows a schematic of the erosion test. The ero-

sion test was conducted using the Sn–3Ag–0.5Cu lead-free

solder melted at 350°C. Erosion behavior is strongly

affected by the flow rate of the molten solder.[2] Gener-

ally, erosion is accelerated in parts such as an impeller, a

rotating shaft or a nozzle, where the flow rate of the molten

solder becomes relatively high. In a previous study, it was

reported that the maximum erosion depth of SUS304 stain-

less steel in molten Sn–3Ag–0.5Cu solder at 350°C increases

with increasing rotation rate of the specimen.[4] In the

study, good accelerating characteristics were observed at

rotation rates ranging from 0 rpm to 100 rpm.[4] Since

higher rotation rates make the test difficult because of

abundant dross formation and splashing of the molten sol-

der, the rotation rate of a rotation axis was 100 rpm in this

study. Dross, which was formed on the surface of the mol-

ten solder, was removed approximately every 12 hours.

When the volume of the molten solder decreased due to

dross formation, a suitable amount of the solder was added

into the solder bath.

After immersion of 50, 100, 200 and 300 hours, the

attached molten solder on the stainless steel surface was

removed with a Teflon brush and the erosion depth was

measured with laser scanning equipment (optWare, Rapid

3D-2000HDS). This equipment can measure erosion

depths at intervals of 5.6 μm from an area approximately

20 × 20 mm in one measurement and convert them into

color mapping data. In this study, twenty points which

have relatively large erosion depths after immersion of 100

hours were evaluated for each specimen using time series

analysis.

An extreme value analysis was also conducted. Except

for the SUS304 specimen immersed for 50 hours, an area

approximately 8 × 33 mm on each side was divided into ten

sub-areas and the maximum erosion depth in each sub-

area was investigated. Since the area at which erosion

occurred was relatively small in the SUS304 specimen

immersed for 50 hours, the extreme value analysis was

conducted for an 8 × 20 mm area on each side. For the

extreme value analysis, the erosion area needs to be

equally divided and each divided area has to include ero-

sion points. Thus, the analysis area in the SUS304 speci-

men immersed for 50 hours was smaller than that of any

other specimen. This difference in the analysis area could

negligibly affect the analysis results. The extreme value

analysis was conducted on both sides of each stainless

steel sample, and thus twenty data points were obtained

for each test condition.

An additional immersion of 2 hours was conducted for

specimens after the immersion test of 300 hours in order

to attach the solder to the surface of stainless steel. For

these specimens, a microstructural observation of the ero-

sion interfaces was conducted using an EPMA.

3. Results and Discussion3.1 Erosion test results

Figures 2 and 3 show general views of the SUS304 and

SUS316 stainless steel specimens after the erosion test.

After immersion of 50 hours, the occurrence of erosion

was observed in both steels. The erosion area spreads with

Fig. 1 Schematic of erosion test.

Fig. 2 General views of SUS304 stainless steel specimenafter erosion test.

37

increasing immersion time.

Figure 4 shows examples of the erosion depth measure-

ment results. In the figure, erosion depth is deeper where

the mapping color is darker. A white area corresponds to

an area at which erosion is negligible. Moreover, black

points correspond to analysis-impossible points using laser

scanning. The color concentration of the mapping result is

relative in every specimen. As shown in Fig. 4, erosion

does not proceed uniformly over the entire specimen. The

erosion depth was investigated from the difference in

depth data between the eroded areas and the non-eroded

areas. In the extreme value analysis, the analyzed area as

shown in Fig. 4 was evenly divided into ten sub-areas.

Each sub-area was repeatedly analyzed using color map-

ping until the maximum erosion depth was recognized in

the area.

Figure 5 shows the relationship between the erosion

depth and immersion time. A linear relationship was found

between the erosion depth and the square root of immer-

sion time in both steels. Their relationships are expressed

as the following equations.

E304 = 0.103 × t 0.5 (1)

E316 = 0.086 × t 0.5 (2)

E304 : Erosion depth of SUS304 steel (μm),

E316 : Erosion depth of SUS316 steel (μm),

t : Immersion time (s)

The erosion speed of SUS304 steel was approximately

20% faster than that of SUS316 steel.

Figure 6 shows the extreme value analysis results. In

the figure, the extreme value and the standard deviation

which were investigated by extreme value analysis using

twenty measurement data points are plotted. A linear rela-

tionship was observed between the maximum erosion

depth and immersion time. Different from the results

shown in Fig. 5, the maximum erosion depth of SUS304

steel is very close to that of SUS316 steel. Therefore, it was

found that the maximum erosion depth does not depend

on steel type under the conditions investigated.

In this study, polyvinyl chloride was attached to the sur-

face of the stainless steel to accelerate the occurrence of

Fig. 3 General views of SUS316 stainless steel specimenafter erosion test.

Fig. 4 Erosion depth measurement results by laser scanningequipment. (Black points correspond to analysis-impossiblepoints.).

Fig. 5 Effect of immersion time on erosion depth of SUS304and SUS316 stainless steels by molten Sn–3Ag–0.5Cu solder. Fig. 6 Extreme value analysis results.

Shohji et al.: Comparison of Erosion Rates of SUS304 and SUS316 Stainless Steels (3/5)

38

Transactions of The Japan Institute of Electronics Packaging Vol. 2, No. 1, 2009

stainless steel erosion by the molten lead-free solder. The

additional experiment revealed that erosion occurs after an

erosion test of 25 hours even in the SUS316 specimen,

which has better corrosion resistance than SUS304. Thus,

the erosion of stainless steel seems to occur in an erosion

test of several hours. From a similar erosion test with the

bare SUS304 specimen, it was reported that erosion of

SUS304 occurs after an erosion test of 300 hours.[5] More-

over, we have conducted similar erosion tests using the

SUS316 specimen without polyvinyl chloride and it was

confirmed that the erosion of SUS316 specimen occurs

after an erosion test of 650 hours. These results show that

the incubation period of erosion occurrence of SUS304 is

shorter than that of SUS316. Since the data shown in Figs.

5 and 6 scarcely include the incubation period, we should

pay attention when we compare the erosion depth of bare

stainless steels with respect to immersion time in the mol-

ten Sn–3Ag–0.5Cu solder.

3.2 Microstructures of erosion interfacesFigure 7 shows the EPMA mapping analysis results of

cross sections of erosion interfaces after an erosion test of

302 hours. The reaction layer, with a thickness of approxi-

mately 30 μm, formed at the interface of SUS304 steel and

the solder. The reaction layer consists of Fe, Cr and Sn.

Analogous layer formation has been reported at the inter-

faces of Sn/Fe, Sn/steel, Sn-based solder/Fe and Sn–Ag–

Cu solder/SUS304 steel [6–8], and the layer was inferred

to be FeSn2 including Cr.[6, 8] On the basis of the results

of the EPMA quantitative analysis, the atomic ratio of Fe,

Cr and Sn in the reaction layer was clarified to be Fe : Cr :

Sn = 29.4 : 3.8 : 66.8. Therefore, the reaction layer was

inferred to be (Fe, Cr)Sn2. Ni diffuses uniformly into the

solder. Coalescence of Si in the solder was also observed.

In the case of SUS316 steel, the reaction layer, with a

thickness of approximately 60 μm, formed at the interface

of SUS316 steel and the solder. The thickness of the reac-

Fig. 7 EPMA analysis results of cross sections of stainless steel/solder interfaces after erosion test for302 hours. (a) SUS304 stainless steel, (b) SUS316 stainless steel.

39

tion layer is double that at the interface of SUS304 steel.

The reaction layer is separated into two layers and both

layers consist of Fe, Cr, Mo and Sn. From the results of the

EPMA quantitative analysis, the atomic ratios of Fe, Cr,

Mo and Sn were evaluated as 18.7 : 7.4 : 2.8 : 71.1 and 26.3 :

2.1 : 2.2 : 69.4 in the reaction layers formed in the vicinity

of the interface and in the solder side, respectively. The

reaction layers include a few at% Mo different from that

formed at the interface of SUS304 steel and the solder. It

seems that Mo forms a thicker reaction layer at the

SUS316 steel/solder interface. It has been reported that

Mo probably exists as a silicide in the reaction layer.[8] A

similar tendency is observed in Fig. 7(b). However, the

effect of Mo on the formation of the reaction layer has not

been clarified yet, and thus further study is required.

In this study, the average erosion rate is slower when

the reaction layer is thicker. For erosion to proceed, ele-

ments of stainless steel should diffuse through the reac-

tion layer formed at the erosion interface and be dissolved

into the molten solder. As shown in Fig. 5, the average ero-

sion depth is in proportion to the square root of immersion

time. When such a relation holds, the erosion process is

mainly controlled by volume diffusion. Therefore, the

progress of the erosion of stainless steel would be mainly

controlled by the volume diffusion of elements of stainless

steel in the reaction layer. The thicker reaction layer

delays the erosion of stainless steel, and thus the average

erosion rate of SUS316 is slower than that of SUS304. How-

ever, the maximum erosion depth evaluated by extreme

value analysis does not depend on steel type as shown in

Fig. 6. The erosion of stainless steel does not proceed uni-

formly as shown in Fig. 4. Thus, a negligible difference by

steel type is observed in maximum erosion depth.

4. ConclusionIn this study, an erosion test was conducted to compare

the erosion rates of SUS304 and SUS316 stainless steels by

molten Sn–3Ag–0.5Cu lead-free solder. The results

obtained are as follows.

(1) The average erosion rate of SUS304 steel is faster

than that of SUS316 steel.

(2) The maximum erosion depth evaluated by extreme

value analysis does not depend on steel type.

(3) An Fe–Cr–Sn layer forms at the erosion interface

of SUS304 steel. On the contrary, an Fe–Cr–Mo–

Sn layer forms at the erosion interface of SUS316

steel. The thickness of the Fe–Cr–Mo–Sn layer

was double that of the Fe–Cr–Sn layer after an ero-

sion test at 350°C for 302 hours.

AcknowledgementThe authors express their gratitude to Ms. Jun Ran and

Mr. Hiroshi Kikuchi (optWare co., ltd.) for their kind assis-

tance in conducting the erosion depth measurement. This

study was conducted as a research activity of Japan Elec-

tronics and Information Technology Industries Association

(JEITA).

References

[1] M. Takeuchi, K. Hayakawa, T. Hosaka, O. Satou, J.

Utumi, M. Yoshikawa and T. Wataya, “Erosion charac-

teristic of solder to the stainless steel and correspon-

dence to the flow tank”, Proc. of MES2002, 147–150

(2002).

[2] T. Takemoto, H. Nishikawa, K. Serizawa and K.

Yamamoto, “Mechanism of damage of flow soldering

bath by molten lead-free solder and its prevention

methods”, Proc. of Mate (Symposium on Microjoining

and Assembly Technology in Electronics) 2009, 383–

386 (2009).

[3] K. Sumiyoshi, I. Shohji, H. Takase and M. Miyazaki,

“Comparison of erosion rates of SUS304 and SUS316

stainless steels by molten lead-free solder”, Proc. of

Mate (Symposium on Microjoining and Assembly

Technology in Electronics) 2009, 395–398 (2009).

[4] H. Nishikawa, S. Kang and T. Takemoto, “Measure-

ment of erosion of stainless steel by molten lead-free

solder using micro-focus X-ray CT system”, Journal of

Japan Welding Society, 27, 214s–218s (2009).

[5] H. Nishikawa, K. Sabase and T. Takemoto, “Effect of

some factor on erosion of stainless steel by molten

lead-free solder”, Proc. of Mate (Symposium on

Microjoining and Assembly Technology in Electron-

ics) 2009, 387–390 (2009).

[6] T. Takemoto and M. Takemoto, “Reaction between

Iron-based alloys and lead-free solders”, Journal of Japan

Institute of Electronics Packaging, 6, 496–502 (2003).

[7] Y. C. Huang, S. W. Chen, W. Gierlotka, C. H. Chang

and J. C. Wu, “Dissolution and interfacial reactions of

Fe in molten Sn–Cu and Sn–Pb solders”, J. Mater. Res,

22, 2924–2929 (2007).

[8] Y. Sejimo, T. Aoyagi, Y. Kariya and T. Kimura, “EPMA

analysis of joining boundary between lead free solder

and stainless steel”, Proc. of Mate (Symposium on

Microjoining and Assembly Technology in Electron-

ics) 2009, 391–394 (2009).

Shohji et al.: Comparison of Erosion Rates of SUS304 and SUS316 Stainless Steels (5/5)