Reading in UTF-8 encoding in Java results in issues at only some places - utf-8

Question

I am reading a UTF_8 encoded file byte by byte in Java Spring Batch and saving the content in mongodb. The reading is delegated to FlatFileItemReader and the encoding is injected into the bean as UTF-8.
What I do see is that only at one or two specific places, the umlaut character is getting stored with an encoding problem in mongodb. I verified that the content in the original file is okay. What is interesting is that everywhere else, the umlaut character in the original file is interpreted correctly and stored correctly in mongodb.
Could you help with this?
<bean id="recordReader" class="org.springframework.batch.item.file.FlatFileItemReader" scope="step">
<property name="resource" value="file:#{jobExecutionContext['inputfile']}" />
<property name="linesToSkip" value = "1" />
<property name= "encoding" value ="UTF-8"/>
<property name="lineMapper">
<bean class="org.springframework.batch.item.file.mapping.DefaultLineMapper" >
<property name = "lineTokenizer">
<bean class = "com.xyz.NoQuoteDelimitedLineTokenizer">
<property name = "strict" value = "false" />
<property name = "names" value = "Col1,Col2,Col3,Col4,Col5,Col6,Col7,Col8" />
<property name="delimiter">
<util:constant static-field="org.springframework.batch.item.file.transform.DelimitedLineTokenizer.DELIMITER_TAB" />
</property>
</bean>
</property>
<property name = "fieldSetMapper">
<bean class = "com.xyz.MyFieldSetMapper" >
<property name = "ctx" value="#{jobExecutionContext['contextBean']}"/>
</bean>
</property>
</bean>
</property>
</bean>
FieldSetMapper
private void validateFields (String id, ConsumerDO ci, FieldSet fieldset_p) throws Exception
{try {
ci.setCaption(fieldset_p.readString("newCaption"));
}
catch (Exception e) {
}
}