A common way of making older airliners more efficient without increasing their wingspan is to add winglets, either as a retrofit for existing aircraft or as a new, wingletted derivative of a preexisting non-wingletted model.
One particular older airliner for which this doesn't work, however, is the DC-9; winglets for the DC-9 have been studied, but never made it into production. One of the main reasons for this (alongside the relative smallness of the benefit that could be gained by wingletting a DC-9, compared to that gained by wingletting other types of airliner, and the increasing smallness of the world's DC-9-classic1 fleet) is that the DC-9's wing structure makes it unusually difficult to safely and successfully winglet:
APB has investigated a retrofit design for the DC-9 family. Those design studies have not been successful in creating a viable business case. Projected block fuel burn reductions of less than 2 percent are offset by substantial modification costs. The limited potential for the DC-9 is a result of the existing wing structure, which hinders installation of a large winglet, as was possible on the Boeing 737 family. Since the DC-9 has been out of production since the early 1980s, the fleet size has shrunk and the fleet has aged, making the business case for a retrofit winglet or wingtip not as attractive as that for the Boeing 737 and 757 families. [Emphasis added.]
What is it about the DC-9's wing that makes it so hard to winglet, compared to (say) a 737?
1: The DC-9-10 through DC-9-50, but not the DC-9-80 or DC-9-90.