There is probably an "invisible" character sitting before the apostrophe... it could have been a leftover from a conversion from X format into Y format.
Can't know specifically what the mystery character is unless you attach an actual sample. I am going to take an educated guess and assume it is some sort of unicode "joining" character, or some sort of no width space.
Explanation:
On your computer, you most likely have font support for nearly every character under the sun... take an easy example, the Greek letters: alpha (α), beta (β), gamma (γ).
In Sigil + Calibre + your web browser, the Greek letters will appear/work fine.
BUT, your ereader can't support every symbol under the sun. They may have decided to only support a subset of characters used in English: A-Z + punctuation + some very commonly used accented characters (á, è, í, ç).
Let us say I tried to use a character OUTSIDE of that subset, (let us say, Ancient Greek or Chinese characters)... the device will not be able to display that character so it will replace it with "unknown character" square, or maybe this symbol (looks like a black diamond with a white '?'):
�
There are a ton of different "spaces" of different widths/usages, for example, this site lists 20 of them:
http://www.cs.tut.fi/~jkorpela/chars/spaces.html
Side Note:
You typically only run into this situation when dealing with accented/rare characters. For example, here are some I ran into during my book conversions, which turn into a box when read on my Nook:
(This was used in a Turkish name) I with a dot over it:
İ
s with a cedilla:
ş
(This is common in Czech) c with a caron:
č
(This was a part of ancient greek quotes) lowercase alpha with a psili and oxia:
ἄ