We propose a combination of a structural path analysis (SPA) with the complex network analysis to capture the existence of different sector or industry clusters in the transmission of carbon emissions from China to the rest of world's economy through imports. We used OECD-ICIO input-output framework and identified different types of communities as boosts and suppliers of carbon emissions by type of industry, type of final demand, and the different stages of production. We have found that some worldwide industries, such as consumption of textiles and wholesale and retail trade, could reduce the emissions generated in China by collaborating with their direct suppliers since they receive a large part of their emissions from the first and second stage of production. However, it is not so easy for other consumption industries, such as food, computers, motor vehicles and most of the service industries, which are more diffuse ones, and incorporate most of their carbon emissions from China from furthest stages of production (3 to 9). In terms of economic policy, this imply that if these industries (or firms) that supply the final demand of the world economy want to reduce a significant part of their emissions generated in China, it should be through establishing collaboration agreements with the suppliers in China that are in the remote stages of the production process, which entails higher transaction costs. (C) 2020 Elsevier B.V. All rights reserved.