Electroless Nickel / Immersion Gold Process Technology for

Improved Ductility of Flex and Rigid-Flex Applications

By: Kuldip Johal and Hugh Roberts - Atotech USA Inc.
Sven Lamprecht and Christian Wunderlich - Atotech Deutschland GmbH
kjohal@[Link]; [Link]@[Link]; hroberts@[Link]

ABSTRACT Figure 1, the flexible printed circuit market is forecasted to

Because of their numerous functional design and grow at an annualised rate of more than 12-percent,
application possibilities, the use of flex and rigid-flex PWBs reaching a worldwide production value of $8.6 billion by
is increasing rapidly. However, this shift poses additional 2008. As indicated in Figure 2, this increase is concentrated
challenges within circuit board fabrication and assembly in key areas, mainly attributed to the rapid growth in use of
operations, particularly in terms of the surface finish on the mobile phone, digital camera and LCD technologies.
PWB. In response, to improve the reliability of flex and
3,000
rigid-flex circuit applications, alternative surface finish

Flex Circuit Production

methods are increasingly being used, such as the electroless 2,500
2003
nickel/immersion gold (ENIG) process. 2,000
2008

($M)
1,500
To avoid cracking of the nickel during bending, the use of
ENIG for flex and rigid-flex circuits has typically required a 1,000
relatively thin electroless nickel deposit of medium 500
phosphorus content. In this technical paper, the reliability
-
of ENIG with high-phosphorus electroless nickel is

er

y
e
r

ns

ry

g
examined for such applications. Using electroless nickel

d
te

la
iv

in
/In
um

ta
io
pu

sp
ot

ag
ili
al
at

m
ns
deposits of varying thickness and phosphorus content,
m

Di
M

ck
ic
ic

to
Co

Co

ed

Pa
un

Au
results of bending tests, hardness, and SEMs are compared

M
m
m
Co

to predict reliability of the assembled flexible circuit. In

particular, Young’s Modulus (stress-to-strain ratio) is Fig 2. Flexible Circuit Market Growth by Application 2003
evaluated to show a direct relationship between the vs. 2008 (Source: Prismark Partners LLC)
phosphorus content of the electroless nickel deposit and the
ductility of the overall ENIG finish. The ENIG process with
high-phosphorus nickel is shown to offer a more reliable Only a few years ago, Japan was recognized as the global
surface finish for flex and rigid-flex applications. center for fabrication of flexible circuits. However, as
expected, much of the growth in this market segment will be
INTRODUCTION experienced most dramatically in other parts of Asia. As
During the next several years, no other segment of the presented in Figure 3, the value of flexible circuits produced
printed wiring board (PWB) industry is forecasted to grow in Asian (countries other than Japan) is predicted to reach
as rapidly as the flexible circuit segment. As shown in nearly five billion dollars by 2008.
($ Bn)
35
Americas
Europe
6,000
30 Japan
Asia (Ex. Japan) 5,000 2003
Flex Circuit Production

25 2008
3.7% 4,000
CAAGR
20
($M)

$17.86
3,000
$14.88
15
12.3%
2,000
1.5%
CAAGR CAAGR
10 $8.59
$7.84 9.9% 12.1% 1,000
$7.27 CAAGR CAAGR
$5.36 $5.35 $4.80
5
M44.044kk-growth

$3.34 $3.03 -
Asia Japan Europe Americas
0
2003 2008 2003 2008 2003 2008 2003 2008 2003 2008
Commodity Multilayers Microvia Si Platform Flex
Fig 3. Flexible Circuit Market Growth by Global Area
Fig 1. Projected PWB Market Profiles Growth 2003 vs. 2003 vs. 2008 (Source: Prismark Partners LLC)
2008 (Source: Prismark Partners LLC)
Figure 4 presents a prediction of the expected flexible SURFACE FINISH ALTERNATIVES FOR FLEXIBLE
CIRCUIT APPLICATIONS
50,000
Currently, a variety of methods exist for surface finish of
Production (m2 X 1000)

flexible and rigid-flex circuits. Among these are:

40,000
2003
• Electrolytic Tin/Lead
30,000 2008 • Electrolytic Tin
• Electrolytic Nickel/Gold
20,000 • Immersion Silver
10,000 • Immersion Tin
• Electroless Nickel/Immersion Gold
-
Single Double Multilayer Rigid/Flex IC
• Organic Solderability Preservative
Sided Sided ML Substrate
There is no single deposit that provides the perfect surface
finish, which explains the existence of these various
Fig 4. Flexible circuit market by type of material alternatives. For example, electrolytic nickel/gold is most
(Source: Prismark Partners LCC) predominant in flexible circuit applications where metallic
surface finishes are used. However, as with any
circuit market growth according to the type of circuit board. electrodeposited metal, there are problems with surface
As shown, single-sided flexible material will continue to be distribution and plating in fine-line dimensions. Likewise,
the dominant format and will increase in terms of market OSPs are simple to use from a fabrication standpoint,
share. although these coatings do not allow wire bonding and their
ability to withstand multiple thermal excursions during
component assembly is well known. As a result, OEMs are
frequently seeking ways to improve the reliability of the
surface finish while reducing costs.

The electroless nickel / immersion gold (ENIG) process has

been used for more than 20 years in the PWB industry. As a
finish, ENIG is now receiving increased attention because it
meets requirements for lead-free assembly while offering a
coplanar surface that is both solderable and aluminum-wire
bondable. ENIG is also well suited for hot bar soldering
and anisotropic conductive film (ACF) bonding.

HIGH-PHOSPHORUS ELECTROLESS NICKEL /

IMMERSION GOLD
In the application of electroless nickel, the nickel is
commonly co-deposited with phosphorus. Most ENIG
processes currently used for circuit applications create a
nickel deposit with medium-phosphorus content, in the
range of 7-9 percent by weight. In recent years there has
been a gradual but consistent shift to the use of high-
phosphorous electroless nickel/immersion gold (HP-ENIG)
as a final finish. This acceptance is particularly evident in
the telecommunications industry, which is experiencing a
significant increase in the use of flexible circuits, as
previously mentioned.

The immersion gold step of any ENIG process relies on the

exchange of nickel ions for gold, which is essentially a
corrosion action. To compensate for the lower ductility of
medium-P electroless nickel deposits, some fabricators of
flexible PWBs finished with ENIG deposit a relatively low
nickel thickness (2-3 microns) in comparison to deposits on
standard rigid materials. This reduced thickness has been
Fig 5. End-use applications for flexible and rigid-flex necessary to avoid nickel cracking during normal bending of
circuitries (Source: Mektec) the flex circuit. However, because of the thin nickel layer,
corrosion from the immersion gold step can frequently
cause problems with solder joint integrity after assembly, a
result commonly known as the “Black Pad” defect. used to measure the gold thickness is also limited in
Although a thicker nickel deposit may eliminate the black accuracy (typically +/- 0.01µm or greater) depending on pad
pad issue, it compounds the problem of nickel cracking size and collimator been used, it does suggest that thicker
during bending of the flexible circuits. This cracking is gold will be deposited on the nickel layer that is more
directly related to the nickel deposit properties, such as readily attacked.
ductility and internal stress, which are primarily influenced
by the composition of the nickel solution, the solution age From the standpoint of solder joint integrity, previous
as defined by number of metal turnovers (MTO) and the investigations1 have shown that the resulting deposit
phosphorous content. exhibits greater ductility and is less prone to brittle fracture
than that obtained using a medium-P process.
HP-ENIG involves the use of an electroless nickel deposit
containing 10-13 percent phosphorus by weight. Because of FLEXIBLE CIRCUIT BENDING TEST METHODS
the higher phosphorus content in the nickel deposit, it offers Because flexible circuit designs are often unique for each
superior corrosion resistance compared to that of a low- or application, the original equipment manufacturer (OEM)
medium- phosphorous process. Figure 6 illustrates the defines their specific performance criteria. However, such
build-up of the electroless nickel and immersion gold layers specifications are typically based on accepted industry
on the base copper of the flexible circuit. standards. The primary standard regarding performance
testing for flexible circuits is IPC-6013A (Qualification and
Immersion Au (0.03-0.05 µm) Performance Specification for Flexible Printed Boards)2.
Within this standard, Section 3.6 specifies “Physical
Electroless Ni (5.0-6.0 µm) Requirements” for such circuits, including “Bending
Flexibility”. Figure 7 shows the basic premise for the 90º
and 180º bend tests according to IPC-6013A, where
Base Copper

Flexible Substrate

Fig 6. Layer build-up for the High-Phosphorus

ENIG process

Table 1 presents information regarding the HP-ENIG

process sequence and key operating parameters.

IPC-6013A-3-09
Table 1. HP-ENIG Process Sequence and
Parameters
Process Treatment Treatment Fig 7. Bending test according to IPC-6013A
Step Time (min) Temp (ºC)
Clean 3-6 35-45 direction of bend (a), degree of bend (b), number of bend
Micro etch 1-2 25-35 cycles and the bend radius (d) are required. Guidelines for
determining the minimum bend radius (d) are set forth in
Acid Dip >3 Ambient IPC-2223-A (Sectional Design Standard for Flexible Printed
Activate 1-3 20-25 Boards). Figure 8 illustrates the performance of the 90º and
Electroless Nickel 20-30 80-90 180º bend tests in practice for a selected bend radius.
Immersion Gold 10-12 80-85

If it is accepted that a corrosion-resistant electroless nickel

layer undergoes less attack by the immersion gold reaction,
the resultant gold thickness will be lower in comparison to a
nickel layer with less corrosion-resistance, providing the
immersion gold solution parameters are equal. This
condition was previously observed1 on a nickel deposit of
8.0-percent phosphorus that achieved a gold thickness of Fig 8. Performing the 90º bend test (left) and 180º bend
0.08µm, while a nickel layer of 11.2-percent phosphorus test (right)
yielded a gold thickness of only 0.05µm. Although the tool
Other dynamic testing is employed within the flex circuit BENDING TEST RESULTS
industry and has been previously documented3. An Wire pull/bend tests were performed on flexible circuits
illustration of each of these tests is shown in Figure 9. with surface finishes of medium-phosphorus and high-
phosphorus ENIG. Using a 2-mm wire, tests were
conducted on both deposits with nickel deposit thicknesses
of 1µm, 3µm and 5µm. In all cases, the gold thickness was
held constant at 0.05µm. All conductor widths were 0.5
mm.

Figure 12 shows the results of the 2-mm wire pull/bend test

as performed on the ENIG deposit of medium phosphorus

Fig 9. Illustration of various bending tests for flexible

circuits: (1) Cyclical Rolling Flex Test, (2) Cyclical
Bend Test, (3) Fatigue Ductility Flex Test and (4)
Collapsing Radius Test
1-µm Ni 3-µm Ni 5-µm Ni
One flexible circuit fabricator has adopted a relatively
simple method for testing the ENIG surface finish on Fig 12. Medium-phosphorus ENIG deposit after 2-mm
flexible substrates. The test involves fixing one end of the wire pull/bend test showing evidence of surface cracks
flexible circuit coupon and pulling a wire of known
diameter (d) through the length of the sample as shown in content. Cracks in the deposit are readily noticeable at a
Figure 10. nickel thickness of 3µm and the defect is exacerbated at the
Nickel Layer 5-µm nickel thickness, as seen in Figure 13, which presents

Flex Material

Force

Tape

Fig 10. Wire pull/bend test for flexible circuits

Fig 13. Cross section of 5-µm medium-phosphorus
ENIG deposit on copper following 2-mm wire pull/bend
By simulating both a bending and “curling” effect, this test
test
was considered to be a more demanding and accurate
measure of the ductility required of the deposit. Figure 11 a sample cross-section that clearly shows the magnitude of
shows the performance of this test in practice using a 2-mm the defect. As noted previously, it is for this reason that
wire to roll the bend through the flexible material. fabricators of flexible circuits will compensate for the lack
of ductility in this type of deposit by reducing the electroless
nickel thickness, Unfortunately, such a reaction increases
the probability of “black pad” effect as a result of corrosion
by the immersion gold.

In contrast, the wire pull/bend test results for the HP-ENIG

deposit resulted in no discernible cracks at any thickness
tested, as shown in Figures 14 and 15. The preferred
minimum electroless nickel deposit of 5-µm can thus be
Fig 11. Securing the flex circuit sample (left) and applied without concern for surface cracking. This is a
performing the 2-mm wire pull/bend test (right) significant advantage of the HP-ENIG deposit since it plays
such a major role in the prevention of copper attack by the 150

immersion gold step (and the resultant Black Pad effect).

Internal Stress (N/mm²)

Medium-P Nickel

Tensile
100

50
High-P Nickel

Compressive
-50

-100

-150
5 6 7 8 9 10 11 12 13 14
Phosphorus Content (wt.%)
1-µm Ni 3-µm Ni 5-µm Ni
16. Internal stress of 5-µm electroless nickel
Fig 14.
Fig 14. High-phosphorus ENIG deposit after 2-mm deposits with varying phosphorus content. content.
degree of phosphorus
wire pull/bend test showing defect-free surface
Figures 17 and 18 show the results of these tests for
different nickel thicknesses and electroless nickel solution
metal turnovers (MTO). Comparing the two charts, it can
be observed that the internal stress of the high-phosphorus
deposit is lower and impacted to a lesser degree by both
metal thickness and MTO.

100
Internal Stress (N/mm )
2

1 MTO
75 5 MTO

50
Fig 15. Cross section of 5-µm high-phosphorus ENIG
deposit on copper following 2-mm wire pull/bend test
25

DUCTILITY AND STRESS ANALYSIS 0

Ductility is a measurement of the extent that a material can 0 1 2 3 4 5 6 7 8
be “plastically” deformed before fracture occurs. It is Nickel Thickness (microns)
commonly expressed as percent elongation (%EL) or
percent reduction in area (%RA): Fig 17. Relationship of internal stress and nickel
thickness for medium-P nickel deposit

100
Internal Stress (N/mm )
2

2 MTO
75
In flexible circuit applications, ductility one is an important 3.3 MTO
property of the ENIG deposit. The key to ductility is to
maintain low internal stress to account for the required 50
bending. As shown in Figure 16, electroless nickel deposits
with medium phosphorus content of (7-9 percent) exhibited
internal stress in the tensile range of approximately 60 25
N/mm². Conversely, the high-phosphorus (10-12 percent)
nickel deposit displayed internal stresses of a compressive 0
nature and of lower values in comparison to the medium- 0 1 2 3 4 5 6 7 8
phosphorus deposit.
Nickel Thickness (microns)
In similar manner, the effect of nickel thickness was Fig18. Relationship of internal stress and nickel
examined for medium- and high-phosphorus deposits. thickness for high-P nickel deposit
 ELASTICITY phosphorus content. As shown, the deposits of higher
For the description of the elastic properties of linear objects phosphorus content exhibit a lower Young’s Modulus,
like wires, rods, columns that are either stretched or indicating a higher degree of elasticity. The Young’s
compressed, a convenient parameter of the material is the Modulus of the HP-ENIG is within the range of that for
Young’s Modulus.4 Young's modulus can be used to electrolytically deposited copper, which is typically 70-110
predict the elongation or compression of an object as long as Gpa.
the stress is less than the yield strength of the material.
Young’s Modulus (E) is defined as the ratio of stress to
strain: HARDNESS
Hardness is defined as the measure of a material’s ability to
withstand indentation. For measurements of microhardness,
the Vickers unit of hardness is often used and results can be
directly related to the strength of the material. The
This ratio can also be expressed as: following table compares the Vickers hardness values for
electroless nickel deposits of varying phosphorus content.
As shown, no clear trend was observed regarding the
relationship between phosphorus content of the nickel and
the hardness of the deposit.

where L0 is the equilibrium length, ∆L is the length change Comparison of Hardness for Electroless Nickel
under the applied stress, F is the force applied, and A is the Deposits of Varying Phosphorus Content
area over which the force is applied. Electroless Nickel Phosphorus Vickers
Process Solution Content Hardness
(%)
Low P 3.2 1026
Medium P 9.5 760
Medium-High P 10.6 1040
High P 12.2 802

CONCLUSIONS
Measured in Pascals or Newtons per square meter (N/m²), Based on the investigations performed in this evaluation, the
Young’s Modulus provides a relatively simple method for following conclusions are offered:
comparing the elastic properties of various materials.
1. OEMs exert a major influence in determining the
As a means of comparing the elasticity of HP-ENIG deposit application requirements for flexible circuits. As such,
versus an ENIG deposit of medium phosphorus content, the design and functionality are often not assigned to a
Young’s modulus for each case was measured. A common industry standard.
Fisherscope H100C was used to measure the resultant stress 2. Varying the content of co-deposited phosphorus
and strain. Figure 19 shows a comparison of Young’s directly influences the inherent stress in the electroless
Modulus values for electroless nickel deposits of varying nickel deposit. The internal stress within the nickel
deposit will shift from the tensile range at medium
phosphorus content (7-9%) to the compressive range at
160
10-13% phosphorus. Furthermore, at the higher
Young's Modulus (GPa)

140
phosphorus content, the internal stress of the nickel is
120
less affected by changes in deposit thickness and
100
plating solution age.
80
3. Increasing the content of co-deposited phosphorus can
60 improve the elasticity of the nickel deposit as measured
40 by Young's Modulus. A more elastic (i.e. lower
20 Young’s Modulus) response can be achieved with the
0 high-phosphorus electroless nickel deposit.
3.2 9.5 10.6 12.2 4. For all thicknesses examined, the HP-ENIG deposit
Phosphorus Content (%) showed superior performance in the 2-mm wire
pull/bend test for flexible circuit applications in
Fig 19. Comparison of Young’s Modulus vs. comparison to a medium-phosphorus deposit.
phosphorus content of different electroless nickel
deposits
In summary, results of internal stress analysis, measurement
of elasticity and practical bending tests indicate that the HP-
ENIG process is well suited for applications involving
flexible and rigid-flex circuitry. Because of its (1) improved
resistance to corrosion from the immersion gold step, (2)
lower and compressive internal stress and (3) higher
ductility, the HP-ENIG deposit was determined to be
capable of withstanding more intensive flexural testing than
conventional ENIG processes with nickel deposits of
medium phosphorus content. Further investigations are
necessary and will be performed to fully assess the impact
of different flexible circuit construction and design on
testing results.

REFERENCES
1. Kuldip Johal , Sven Lamprecht, Dr. Hans-Jurgen
Schreier, Hugh Roberts; “Impacts of Bulk Phosphorous
Content of Electroless Nickel Layers to Solder Joint
Integrity and their Use as Gold- and Aluminum-Wire Bond
Surfaces”, Proceedings of SMTA Pan Pacific
Microelectronics Symposium; February 2004.

2. “Qualification and Performance Specification for Flexible

Printed Boards” IPC-6013A, Published by IPC, November
2003.

3. Fjelstad, Joseph. “Bend Testing Methods for Flexible

Circuits”. CircuiTree; February 2002.

4 . Halliday, Resnick, Walker, Fundamentals of Physics,

5E,Extended, Wiley 1997.