SMTAI-PAPER-2020-Richter_Final-2.pdf - 第2页

deposit on the heat exchanger and co oled areas of the heat exchanger b ox and then drip do wn to catch tra ys. T hese systems are typically abo ut 30% efficient in flu x capture and require a hi gh le vel of maintenance…

HIGH-EFFICIENCY, LOW MAINTENANCE FLUX REMOVAL METHOD FOR

REFLOW FURNACES

Paul J. Richter

BTU International, Inc.

23 Esquire Road North Billerica, MA 01862

prichter@btu.com

ABSTRACT

The accumulation of residual flux in solder reflow furnaces

is known to result in product yield losses and maintenance

down time. This is a significant industry problem. Various

techniques are currently used to trap and remove flux

residues from the furnace environment, but the efficiency

and maintenance burden of these methods is not optimal.

Described in this work, is a novel method of capture which

separates flux from the recirculating furnace gas using an

aqueous gas/liquid sequestration process. This system

operates automatically and does not require production line

shut-down for routine maintenance which consists of

periodically emptying a small liquid waste container. This

method is compatible with both Clean and No-Clean type

fluxes. Capture efficiency of different fluxes were tested

and will be reported in this study.

Key words: Reflow Soldering, Flux Management,

Convection Oven

BACKGROUND

The conveyor oven reflow soldering process is now well

established in industry for joining SMT components to

printed circuit boards. Fluxes formulated into solder pastes

to prevent oxidation and aid wetting during the reflow

process produce sticky residues in the oven which, if not

controlled, can drip onto product and accumulate in the

oven. This causes yield losses, equipment down time and

high maintenance costs. A method to remove flux effluents

from process oven gas that is effective, cost efficient and

does not require high maintenance demands has been

elusive even though numerous approaches have been tried

through the years. A new technology for addressing flux

management is discussed in this paper.

COMPOSITION OF FLUX EFFLUENTS

No-clean flux composions now in commercial use typically

have low solids content (~5% by weight). Consequently, a

large percentage of the flux constituents are availible to

vaporise and burn in the oven causing heavy effluent

deposition in the oven and dripping onto product. To gain

information about the composition of the resulting flux

effluent, we reached out to our industry partners and

received samples of flux effluents taken from their

production ovens. A suite of analytical techniques including

ICP-MS, FTIR, GC and XPS was used to characterize the

residue samples. Summary results showed the primary

organic components of the effluent are hydrocarbons with

molecular weights of 200-249 (55-85%) and 250-300 (15-

45%). Lighter hydrocarbons (200-249) generally evolve

earlier in the oven thermal profile, while heavier species

come out later in the profile. The hydrocarbons contain

oxygen (and to a lesser extent nitrogen). The primary

surface structures are C-C, C-H single and double bonds;

however, some O-H, C-O single and double bonds are also

present. The presence of metals was found in the samples

with the largest contributors being: Copper, Potassium,

Silicon, Tin, Zinc, and Iron (see Table 1). The values varied

depending upon sampling location in the oven.

Table 1. Flux Effluent ICP-MS Summary Data

Element

Concentration

(ppm)

14000

1200

629

584

375

315

196

6.6

3.4

2.4

1.1

METHODS OF FLUX REMEDIATION

Condensation flux management is very common and is

currently the standard of the industry. The condensation

method takes flux laden process gas and generally passes it

through a disposable, or cleanable pre-filter element and

then passes it through a heat exchanger cooled by air, or

water. This cools the gas and allows flux to condense and

deposit on the heat exchanger and cooled areas of the heat

exchanger box and then drip down to catch trays. These

systems are typically about 30% efficient in flux capture and

require a high level of maintenance to keep the heat

exchangers clean and unclogged of flux.

Catalytic oxidation is also currently being used in different

variations. These systems use a catalyst to promote

oxidation of organic compounds to carbon dioxide and

water by increasing the kinetic rate. The catalyst allows the

oxidation reaction to occur at relatively low temperatures of

about 340 to 500°C. As seen in the analysis of flux

effluents, the effluent product contains high and low

molecular weight hydrocarbon compounds and also a

significant amount of metals. Catalytic oxidizers can

effectively dissociate the low molecular weight

hydrocarbons into carbon dioxide and water, but are

ineffective at cracking high molecular weight resinous

species. Catalyst life is generally short due to catalyst

poisoning. Catalyst poisoning occurs when a compound

bonds to the catalyst’s active sites and chemically

deactivates the catalyst function due to the presence of

metals such as : copper, potassium, zinc, tin and iron.

Sometimes methods are used to pre-clean the gas before it

reaches the catalyst to try to extend the life of expensive

catalysts. Also, regeneration cycles can be used to

reactivate the catalysts. Resinous, metal laden flux effluents

are very difficult to clean out of the gas stream with

catalytic methods.

GAS-TO-LIQUID SEQUESTRATION

This is a new flux remediation technology that was

developed with the goal of providing a high-efficiency flux

capture method which requires minimal maintenance and

handling of flux residues. This system uses a dynamic

liquid filter to separate flux effluents from the oven gas

stream and suspend them in a liquid medium.

The liquid solution which acts as the transfer medium for

this system is water based, safe to handle, non-flammable

and low cost. The solution is comprised of water that is

provided to the system by plumbing directly to the plant

potable water supply and a concentrated detergent which is

automatically added to the water in the proper ratio.

The gas-to-liquid sequestration unit functions by taking

effluent gas from the oven and passing it through a dynamic

liquid filter where the effluent in the gas stream interacts

with the moving liquid and is entrained and suspended.The

cleaned gas then passes through a condenser to remove

latent moisture from the cleaning process. The gas is then

recycled back to the oven process chamber (see Figure 1).

During operation, a small portion of the solution is diverted

to the waste carboy and then the level of solution is made

back up with an automatic addition of water and

concentrate.

Figure 1. Gas-To-Liquid Sequestration Diagram

RESULTS OF EXPERIMENTS

This data includes a determination of flux capture

efficiency, the ability of the gas-to-liquid method to capture

metals present in the process gas stream and a test to see if

this new process has any effect on solder wetting.

To investigate flux capture efficiency, a laboratory

apparatus was fabricated to simulate oven process gas

operation and introduce a precise amount of burnt and

vaporized flux effluent into the gas-to-liquid unit (see

Figure 2). With this closed circuit apparatus, the gas-to-

liquid sequestration unit was tested independent of any

influence that the oven might contribute. This apparatus can

be run in either nitrogen, or air atmosphere. For this

experiment, nitrogen was chosen to allow testing the system

for potential effect on oven oxygen level.

Figure 2. Test Apparatus Diagram

GAS

OUT

Two commercial no-clean fluxes were run in this

experimental series. One flux was a high-tack formulation,

the other was a standard formulation. The gas-to-liquid

removal unit was run in two different configurations for this

testing. Configuration #1 was run without using the

drain/replenishment function for the liquid solution. The

reasion for running this configuration was to determine the

inherent capacity of the liquid solution to uptake flux. The

second configuration was run the same as the first except

with the automatic drain/replenishment function activated.

The reason for running this configuration was to test

efficiency during continous automatic operation. For this

series, process gas temperature was maintained at 320°C.

and flux was processed in 200 gram increments. The

capture efficiency was determined by separating captured

flux from the liquid solution and weighing it. The weight of

the captured flux was then expressed as a percentage of the

flux introduced into the test apparatus. Condition 1, running

with a fixed quantity of liquid solution, maintained stable

efficiencies of ~94% until 1600 grams of flux had been

injected into the apparatus (see Figure 3). As more flux was

introduced, a steady decline in efficiency (though still at

high efficiencies) occured as the solution began to saturate.

This result demonstrates a large inherent capacity of the

liquid solution for capturing and holding flux effluents.

Also of note is that both flux types tested performed

essentially the same.

Figure 3. Condition 1 Flux Capture Efficiency

Condition 2, was run in normal automatic mode and the

system was set to perform a partial drain/replenishment

cycle for each 200 gram addition of flux. 5 kilograms of

each flux type were processed in this test. The capture

efficiency for this run was determined by separating the flux

from the waste liquid released in the drain cycle and

normalizing for sample size, comparing it to the cumulative

input flux and again expressing as a percentage of input.

Results from this condition show continuous stable running

efficiencies of between 90 and 95% (see Figure 4). Again,

both fluxes performed essentially the same.

Figure 4. Condition 2 Flux Capture Efficiency

During these efficiency tests which were run in full

nitrogen, the oxygen level was monitored at both the gas

input and output ports of the flux sequestration unit. For

these test runs, oxygen levels were stable and measured at

between 40 – 48ppm (see Figure 5). The oxygen levels

were not negatively affected by the flux capture process.

Figure 5. Inlet and Outlet Oxygen Levels

To investigate the effectiveness of the gas-to-liquid

sequestration method at trapping the metals found in solder

flux effluents, a sample of liquid solution waste from a

sequestration unit running on a commercial reflow line was

prepared for ICP analysis. The results were consistent with

those seen on samples taken directly from the inside of an

oven (see Table 2). This was an interesting result which

indicates a capacity of the new technology to trap metals.

Future work is planned to quantify metals capture by using a

100

-0.5 0 0.5 1 1.5 2 2.5

Flux #1 (Hi-Tack)

Flux #2

Flux Capture Efficiency (%)

Cumulative Flux Added (Kg.)

100

-1 0 1 2 3 4 5 6

Flux #1 (Hi-Tack)

Flux #2

Flux Capture Efficiency (%)

Cumulative Flux Added (Kg.)

-1 0 1 2 3 4 5 6

Outlet

Inlet

Oxygen (ppm)

Cumulative Flux Added (Kg.)