Social media is image driven. It plays on the most basic sense we have. Sight. Humans are geared towards information that is seen through our eyes because a lot can be conveyed quickly through even just a blink. Facebook and definitely Instagram have been shaped by their creators to push to the top posts whose images/videos are engaging. Facebook has deemed that images with text on them, even just a little bit, are not eye-catching enough and will not give them much love in the newsfeed. This makes sense. Our eyes want what they want. Pure, unadulterated, eye candy. Text equals work, and we know how we feel about work.
Leave the text for the body of the post. Use emojis when it makes sense to make your text fun and easier for the brain to be like "okay... I see a few cartoons... I guess I'll read the words in between." Emojis are like the ranch dressing to your veggie tray. They make your important but dull words palatable. Just be careful, no one wants to drink ranch through a straw (well, I am sure there are a few of us who would). Emojis can turn a post into a boring felt board in a hurry. Felt boards were not bad, but our audiences are not five either. Moral of the story. No text on an image on social media. Ever.